evipedia.ai
← Back

Trehalose for Health & Longevity

Evidence Review created on 11/01/2026 using AI4L / Opus 5.5

Also known as: α,α-Trehalose, D-(+)-Trehalose, Trehalose Dihydrate, Mycose, Tremalose, Treha, SLS-005

Motivation

Trehalose is a sugar made of two glucose molecules joined by an unusually tough link. Mushrooms, yeast, insects and desert plants make it to survive drying out, freezing and heat. Humans do not make it, but an enzyme in the small intestine splits it into ordinary glucose. Since the late 1990s an enzymatic process has made it cheap enough to use widely as a food ingredient, and it now appears in baked goods, frozen desserts, eye drops and injectable medicines, where it keeps proteins from falling apart.

What drew longevity researchers to it is a laboratory question: whether trehalose can switch on, in cells and in animals, the housekeeping system that clears away damaged proteins, a system that becomes sluggish with age. That question prompted work in models of brain disease and metabolic disease, and later in people.

This review examines what is known about trehalose in humans: how much of a swallowed dose survives digestion, which effects have been tested in controlled trials and which rest only on cells and animals, what doses were used and by which route, what the digestive and other tolerance limits are, and where the published evidence is contested.

Benefits - Risks - Protocol - Conclusion

High-level sources that give an overview of trehalose in human health rather than reporting a single trial.

Only three items are listed because no further overview-level source qualified; a short commentary on a single 2016 study was excluded for the same reason as single-study write-ups. No content on trehalose was found from Rhonda Patrick, Peter Attia, Andrew Huberman or Life Extension Magazine; the single Chris Kresser hit mentions trehalose only in passing, and Lifespan.io’s trehalose articles each report a single study rather than giving an overview, so both were excluded.

Grokipedia

  • Trehalose

    Covers chemistry, natural occurrence, the TFEB-driven autophagy mechanism (cellular recycling of damaged components; TFEB is the master switch for autophagy genes), intestinal metabolism, regulatory status, and industrial and medical uses.

Examine

  • Trehalose

    States the central practical problem plainly: trehalose appears to have therapeutic cellular mechanisms, but poor absorption of intact trehalose and rapid digestion may preclude any benefit from swallowing it.

ConsumerLab

No publicly accessible ConsumerLab article on trehalose exists: its only coverage is a members-only trehalose section within its sugar-substitutes answer, so no link is given.

Systematic Reviews

Systematic reviews and meta-analyses of trehalose located on PubMed, selected for relevance, size and recency.

Trehalose involves a trade-off between its proposed metabolic and protein-maintenance benefits and its gastrointestinal and microbial risks. The benefit side is represented above; no systematic review or meta-analysis covers the principal risks — dose-dependent gastrointestinal intolerance, or whether trehalose favours Clostridioides difficile (a gut bacterium that causes severe, often antibiotic-related diarrhoea) — so the risk side of the trade-off is unrepresented in this literature.

Mechanism of Action

Trehalose is two glucose units joined by an unusually stable α,α-1,1 bond. In humans it is cut by trehalase (the intestinal enzyme that splits trehalose), which sits in the lining of the small intestine, so, per a review, most of an oral dose reaches the blood as glucose (Chen 2023). Cytochrome P450 enzymes (the main drug-metabolising enzymes) are not involved.

Where intact trehalose does reach cells, two mechanisms are proposed. It blocks the SLC2A family (GLUT, the transporters that carry glucose into cells), which starves the cell of glucose and activates AMPK (AMP-activated protein kinase, the cell’s low-energy sensor), triggering autophagy — the recycling of damaged proteins and organelles (DeBosch 2016); that study also saw reduced signalling from mTOR (mechanistic target of rapamycin, the main growth-signalling hub), although trehalose was first described as acting independently of it (Sarkar 2007). It also accumulates inside lysosomes (the cell’s disposal compartments), raises their pH slightly and releases TFEB (transcription factor EB) into the nucleus (Jeong 2021).

A competing reading holds that trehalose blocks autophagic traffic rather than inducing it: in human and rat neuronal cultures it raised levels of autophagosomes (vesicles that collect cargo for disposal) and p62 (a cargo-tagging protein that builds up when disposal stalls) while reducing autolysosomes (the vesicles where cargo is digested) (Yoon 2017). Separately, trehalose binds expanded polyglutamine stretches (abnormally long protein repeats) directly and stabilises partly unfolded protein, a chaperone-like action independent of autophagy (Tanaka 2004). No published human study reports its half-life, tissue distribution or clearance.

Historical Context & Evolution

Trehalose was isolated in 1832 from ergot of rye and later from trehala manna, the cocoon of a weevil, which gave it its name. For 150 years it was a curiosity: the molecule that lets yeast, insects and resurrection plants survive desiccation. Its original practical use was industrial, as a protein and membrane stabiliser. That remains its largest use, in freeze-dried vaccines, antibody formulations and frozen foods.

An enzymatic process developed by Hayashibara in the 1990s cut the cost from a specialty reagent to a commodity sweetener. United States food use followed when the FDA closed its review of GRAS Notice GRN 45 (generally recognized as safe) in October 2000 with no questions, and the European Union authorised it as a novel food in Commission Decision 2001/721/EC.

Health interest came from two findings. Oral trehalose reduced polyglutamine aggregates and extended lifespan in a mouse model of Huntington disease (an inherited brain disease) (Tanaka 2004), and trehalose was then identified as an autophagy enhancer that does not act through mTOR (Sarkar 2007). A decade of animal work followed in arterial ageing, fatty liver and atherosclerosis.

Two counter-currents shape the present. Collins 2018 proposed that food-additive trehalose helped epidemic Clostridioides difficile lineages spread; a human genomic cohort found no association with patient outcomes, and dietary exposure was recalculated as small (Eyre 2019, gut-model work funded by Hayashibara). And because trehalose is unpatentable, trial funding has gravitated to patentable forms — an injectable formulation and trehalase-resistant analogues — rather than to large food-ingredient trials.

Expected Benefits

High 🟩 🟩 🟩

Lower After-Meal Glucose and Insulin Response When Trehalose Replaces Glucose

Trehalase must split trehalose before its glucose reaches the blood, so the same carbohydrate load produces a flatter glucose and insulin curve than the same load of glucose does. Crossover trials from two independent groups show this: van Can and Blaak at Maastricht University, in overweight adults with impaired glucose tolerance (a blood sugar response above normal but below the diabetes threshold), and Yoshizane and colleagues at Hayashibara, which manufactures and sells trehalose, in healthy volunteers. It is a substitution effect only; trehalose still supplies four calories per gram.

Magnitude: Co-ingested with a mixed meal, trehalose attenuated the rise in glucose by 33% and in insulin by 14% compared with an equal load of glucose in the same ten adults with impaired glucose tolerance (van Can 2012); the paper reports no 95% confidence interval (CI, the range in which the true value probably lies). A 25 g dose produced no rapid glucose rise and lower insulin and gastric inhibitory polypeptide (a gut hormone that promotes fat storage) than 25 g of glucose in twenty volunteers (Yoshizane 2017).

Improved Exercise Performance in the Later Stages of Prolonged Exercise

Taken as a drink during endurance exercise, trehalose improved late-stage work output over water. Randomized crossover trials from two independent groups show this: de Oliveira and Saunders in São Paulo, in thirteen trained male cyclists, and Hamada and Wadazumi in Osaka, whose trials carry a co-author from Hayashibara, which manufactures and sells trehalose. Blood glucose stayed flatter than with maltodextrin (a fast-digested starch carbohydrate). No superiority over maltodextrin or isomaltulose (a slow-digested sucrose relative) was shown, and all participants were men.

Magnitude: 302 ± 39 kJ of total work in the 20-minute time trial versus 287 ± 48 kJ with placebo (p = 0.01; p is the probability of a difference this large arising by chance alone); the paper reports no 95% confidence interval (de Oliveira 2024). In the replicating trial, mean power in the final sprint set was 80.5 ± 6.8% on trehalose versus 76.9 ± 7.0% on water (p = 0.017), with no 95% confidence interval reported (Hamada 2021); the concentration-comparison trial gives no between-condition figure (Hamada 2022).

Medium 🟩 🟩

Faster Radiographic Bone Healing and Better Hip Function After Hip Fracture

In a double-blind randomized trial, 3.3 g/day for twelve weeks after surgical fixation of a pertrochanteric fracture (a break through the upper thigh bone below the ball of the hip) improved radiographic union scores and hip function versus sucrose placebo. The proposed basis is the same cytoprotective and autophagy-related action seen in preclinical fracture models. This is a single trial of 64 patients aged 18–60 from one research network, and the authors state adverse events were not systematically monitored, so tolerability at this dose is not established by it.

Magnitude: 2.09 points more than placebo on the Radiographic Union Score for Hip (95% CI 0.45 to 3.72) and 8.55 points more on the Harris Hip Score (a clinician-rated hip function scale; 95% CI 3.97 to 13.13), at week 12 (Mehrvar 2026). Pain fell in both arms with no between-group difference.

Comparable Periodontal Outcomes as a Subgingival Air-Polishing Powder ⭕️ Not Central to Health & Longevity

Trehalose powder is used by dental clinicians in air-polishing devices to clean below the gum line. A meta-analysis of nine randomized trials found it comparable to glycine and erythritol powders; its trehalose arms come from two split-mouth trials by the Kruse group in Freiburg, each comparable to sonic scaling (Kruse 2019, Kruse 2022). This bears on professional periodontal treatment, not on taking trehalose, and nothing about it transfers to oral or intravenous use.

Magnitude: No statistically significant difference versus glycine or erythritol for probing depth, clinical attachment level or bleeding on probing across nine trials in 462 patients; for probing depth, the network analysis ranked trehalose second of three (surface-under-the-curve ranking, a 0–100 score of how likely each powder is to be best, 48.0, versus 84.1 for erythritol and 28.5 for glycine) (Yuan 2025).

Low 🟩

Relief of Dry Eye Disease Symptoms with Topical Trehalose ⚠️ Conflicted

Trehalose–hyaluronic acid drops outperformed comparators on symptoms and tear-film stability across ten trials; three were Théa-funded. Trehalose alone improved signs in one Okayama group’s small trials (Matsuo 2002). The largest trial found no symptom difference (Morya 2021). Net: combination drops help; trehalose’s own share rests on small trials.

Magnitude: Ocular Surface Disease Index score (a validated dry-eye symptom questionnaire) improved by 8.5 ± 7 points more than control drops and tear-film break-up time by 1.9 ± 1 seconds more, averaged across the six and seven trials reporting each (Ballesteros-Sánchez 2023); the review reports ranges rather than 95% confidence intervals. Included trials came from several research groups, such as Schmidl 2015, Chiambaretta 2017 (Théa-supported) and Cagini 2021, all testing trehalose–hyaluronic acid drops; a 0.3% hyaluronic gel gave longer comfort than a trehalose–hyaluronic gel (Karaca 2020).

Improved Oral Glucose Tolerance After Sustained Daily Intake ⚠️ Conflicted

Two randomized trials at the manufacturer Hayashibara reported better glucose tolerance after twelve weeks at 10 g/day (Mizote 2016) and 3.3 g/day, but only against baseline or in an after-the-fact subset. An independent trial in type 2 diabetes found no glycaemic change. Net: an unreplicated, subset-only signal.

Magnitude: At 3.3 g/day, in the subset whose after-meal glucose sat at the upper end of normal, week-12 two-hour glucose was 95 ± 16 mg/dL on trehalose versus 108 ± 14 mg/dL on sucrose (13 per arm, p < 0.01), with no between-group difference in the full sample (Yoshizane 2020). The contrary trial reported between-group differences of −7.1 ± 10.7 mg/dL for fasting glucose and −0.1 ± 0.4% for HbA1c (average blood sugar over about three months), neither significant (Hashemian 2023).

Reduced C-Reactive Protein in Type 2 Diabetes ⚠️ Conflicted

A randomized trial of 3.3 g/day in forty patients with type 2 diabetes reported lower high-sensitivity C-reactive protein (CRP, a blood marker of body-wide inflammation). A randomized trial of intravenous trehalose found no reduction in arterial wall inflammation on imaging (Jamialahmadi 2022). Net: a single-marker signal unmatched in inflamed tissue.

Magnitude: CRP was 0.62 ± 0.3 mg/L lower with trehalose than with sucrose placebo (p = 0.02) at twelve weeks (Hashemian 2023); the paper reports no 95% confidence interval.

Improved Resistance-Artery Endothelial Function in Middle-Aged and Older Adults

In 32 healthy adults aged 50–77, 100 g/day for twelve weeks was compared with maltose on endothelial (blood-vessel lining) function in flow-regulating arteries. The all-participant comparison was null; the gain appeared only in the weight-stable subset, and large-artery function and stiffness did not change. Blocking nitric oxide production abolished it.

Magnitude: Forearm blood flow to acetylcholine rose about 30% within the trehalose arm of the weight-stable subset (13.3 ± 1.0 versus 10.5 ± 1.1 area under the curve, the total response summed across all infusion doses, p = 0.02) while the maltose arm did not change (p = 0.40, no maltose figures reported); the test of whether the two arms changed differently in this subset gave p = 0.04, with no between-group difference figure reported, and across all participants the groups did not differ (p = 0.17) (Kaplon 2016).

Stabilised Ataxia Scores in Spinocerebellar Ataxia Type 3 ⚠️ Conflicted

Weekly intravenous trehalose kept ataxia scores stable over 6–12 months in an open-label series of 14 patients with spinocerebellar ataxia type 3 (a hereditary brain disease). A small, prematurely stopped randomized trial of oral trehalose found no difference. Net: an uncontrolled intravenous signal the oral trial did not reproduce.

Magnitude: The average Scale for the Assessment and Rating of Ataxia score of all 14 patients remained stable at 6 months, as did that of the six treated to 12 months (no control group), with no change figure in the abstract (Zaltzman 2020; the regimen was rated Level B in a systematic review, Yap 2022); 100 g/day oral trehalose showed no difference from maltose in ataxia severity among 15 completers (Yap 2024).

Speculative 🟨

Enhanced Autophagy and Clearance of Misfolded Proteins ⚠️ Conflicted

Basis is cell and animal work; no human study has measured autophagy on trehalose. Intravenous trehalose did not slow amyotrophic lateral sclerosis (a fatal disease of motor nerves) (HEALEY 2025). Net: unverified in people.

Lifespan Extension in Model Organisms ⚠️ Conflicted

Trehalose extended mean lifespan by over 30% in the nematode Caenorhabditis elegans (Honda 2010), but high concentrations shortened female fruit-fly lifespan by 12.5% (Xu 2023). Basis is animal only. Net: inconsistent across species.

Benefit-Modifying Factors

  • Intestinal trehalase activity: The reduced-activity variant rs2276064 of TREH (the gene for trehalase, the enzyme that splits trehalose) lowers enzyme activity. Carriers digest less of a dose, which should blunt any glucose-derived effect while increasing the unabsorbed fraction reaching the colon (Kozlov 2023).

  • Baseline after-meal glucose: The clearest glucose signal appeared in people whose two-hour glucose was at the upper end of normal; the full sample showed no between-group difference, and lower-baseline participants were not analysed separately (Yoshizane 2020).

  • Body weight stability: Weight gain during supplementation erased the vascular effect. The arterial finding held only in participants whose body mass changed by under 2.3 kg (Kaplon 2016).

  • Pre-existing health conditions: In established type 2 diabetes, twelve weeks at 3.3 g/day produced no glycaemic change; no trial has compared responses across metabolic states directly (Hashemian 2023).

  • Sex: Not established. The exercise trials enrolled men only; the vascular, glycaemic and fracture trials enrolled both sexes but reported no analysis by sex, so any difference is unmeasured.

  • Age: The only trial in adults 50–77 used 100 g/day and was null on its randomized comparison (Kaplon 2016); the low-dose trials ran in younger and middle-aged adults, so the dose that works at the older end of the range is untested.

Potential Risks & Side Effects

High 🟥 🟥 🟥

Dose-Dependent Gastrointestinal Intolerance

A double sugar not fully split in the small intestine draws water into the bowel and is fermented, producing bloating, flatulence and loose stools. Two independent groups show this: the Seals group at Colorado recorded these effects at 100 g/day and lost participants to them, and the Oku group in Tokyo found diarrhoea rising with dose in healthy women (Oku 1998). A case report (Bergoz 1971) describes the same symptoms where trehalase activity is low. Symptoms are dose-related and reverse on stopping.

Magnitude: Four of 18 participants randomized to 100 g/day reported bloating, flatulence or loose stools, versus none of 19 on an equal load of maltose, and three withdrew for side effects versus one on maltose (Kaplon 2016). In a dose-escalation study of 20 healthy women, half had no diarrhoea even at 60 g of trehalose, versus diarrhoea in 75% on up to 40 g of the non-digestible comparator lactulose, giving an estimated laxative threshold of 0.65 g/kg body weight (Oku 1998). Single doses up to 50 g were tolerated in the human studies compiled in the safety dossier (Richards 2002). People with low trehalase activity sit at the susceptible end: trehalase deficiency was found in at least 8% of 97 adult Greenlanders biopsied, with no blood-glucose rise after a trehalose load in the three tested (no control group) (Gudmand-Høyer 1988).

Medium 🟥 🟥

Serious Adverse Events with High-Dose Weekly Intravenous Trehalose ⚠️ Conflicted

The only large randomized test of intravenous trehalose recorded more serious adverse events, more discontinuations and more fatal events on drug than on placebo, in adults with amyotrophic lateral sclerosis (ALS, a motor-neuron disease). No death was judged drug-related, and respiratory failure, the commonest cause, matches the natural course of ALS. A smaller randomized trial after myocardial infarction (heart attack) reported an optimal safety profile, but enrolled only 15 patients — too few to detect events at these rates. Net: the imbalance is real but confounded by the underlying disease.

Magnitude: Serious adverse events in 19 of 120 (16%) on trehalose 0.75 g/kg weekly versus 3 of 41 (7%) on regimen placebo, with premature discontinuation in 14 (12%) versus 1 (2%) and seven fatal treatment-emergent events versus none (HEALEY 2025); no excess was seen in the 15-patient coronary trial (Jamialahmadi 2022).

Low 🟥

Caloric and Glycemic Load at Supplement-Level Doses

Trehalose is a caloric sugar, four calories per gram, that yields two glucose molecules. At the doses used to chase autophagy effects it adds to daily energy intake, and the vascular trial at that dose saw weight gain in both arms — which erased the vascular benefit it was testing.

Magnitude: 100 g/day supplies roughly 400 kcal; body mass rose about 0.7 kg by week 4 in both the trehalose and maltose arms (p = 0.02 and p = 0.04 within arms) and did not differ between them (p = 0.87), with individual changes from −2.6 to +3.6 kg (Kaplon 2016).

Promotion of Hypervirulent Clostridioides difficile ⚠️ Conflicted

Two epidemic Clostridioides difficile lineages gained the ability to use very low trehalose concentrations, and dietary trehalose raised the virulence of one in mice. Human data do not support the inference: the utilisation genes are widespread and unassociated with patient mortality. Net: a mouse mechanism not borne out in people.

Magnitude: Dietary trehalose increased the virulence of ribotype 027 (a strain type defined by genetic fingerprint) in a mouse infection model (Collins 2018); among 208 human infections with a lineage variably carrying the four-gene cluster, 30-day all-cause mortality was not associated with it (odds ratio, the odds of death with the cluster relative to without it, 0.36, 95% CI 0.09 to 1.34) (Eyre 2019 — whose gut-model experiments were funded by Hayashibara, the trehalose manufacturer).

Speculative 🟨

Autophagic Flux Blockade and Increased Protein Aggregation

In neuronal cultures trehalose raised α-synuclein aggregation and its release between cells while blocking autophagic traffic (Yoon 2017). Basis is in-vitro only; cell viability was unaffected and no human counterpart exists.

Risk-Modifying Factors

  • Intestinal trehalase activity: The TREH rs2276064 A allele reduces enzyme activity and marks risk of trehalase deficiency. Frequency ranges from 3% in Eastern Slavic to 63% in Chukchi and Koryak populations (Kozlov 2023).

  • Recent antibiotic exposure: Antibiotic-induced disruption of gut bacteria is the precondition for Clostridioides difficile expansion in the animal models, though in antibiotic-treated mice given a hypervirulent strain trehalose did not increase it (Zhang 2020, with a Hayashibara co-author). Inferred from mechanism; no human data.

  • Baseline liver enzymes: Alanine aminotransferase (ALT, a liver enzyme that rises when liver cells are injured) changed differently on intravenous trehalose than on placebo in a 15-patient trial, so an abnormal baseline value makes later changes harder to interpret (Jamialahmadi 2022).

  • Pre-existing bowel disease: A review finds sorbitol absorption depends on absorptive surface and bacterial overgrowth ferments sugars into the same symptoms, yet calls trehalose malabsorption itself clinically unproven (Montalto 2013). Extension to trehalose is inferred; untested.

  • Sex: Not established. No trial reports adverse events split by sex; the gastrointestinal findings come from mixed-sex groups too small for such an analysis.

  • Age: The highest rate of reported side effects came from the oldest cohort studied, adults 50–77 on 100 g/day, but dose and age are confounded in that trial (Kaplon 2016).

Key Interactions & Contraindications

Documented interactions:

  • None identified. No product label, regulator or human interaction or pharmacokinetic study of trehalose exists. Intravenous trehalose was co-administered with riluzole and edaravone (HEALEY 2025) and with post-infarction treatment (Jamialahmadi 2022), but neither trial tested for an interaction.

Theoretical interactions (inferred from mechanism, not studied in humans):

  • Alpha-glucosidase inhibitors (diabetes drugs that block the gut enzymes splitting sugars: acarbose, miglitol): Less trehalose would be digested, increasing the osmotic and fermentable load reaching the colon, with worse bloating and diarrhoea. Keeping trehalose doses small limits it.

  • Metformin: Shares the AMPK activation that trehalose is proposed to produce, and independently causes diarrhoea. Consequence is additive gastrointestinal upset; separating doses by several hours and introducing trehalose at the low end are the usual mitigations.

  • Insulin and sulfonylureas (oral drugs that make the pancreas release insulin: glimepiride, glipizide, gliclazide): Trehalose still delivers glucose. Consequence is higher-than-expected blood glucose when it is treated as a non-caloric sweetener; counting it as carbohydrate when matching insulin doses avoids it.

  • GLP-1 receptor agonists (drugs mimicking a gut hormone that curbs appetite: semaglutide, liraglutide, dulaglutide): Both slow gastric emptying and provoke nausea. Consequence is additive nausea, fullness and diarrhoea; separating the trehalose dose from the injection day is the usual mitigation.

  • Osmotic laxatives and sugar alcohols (lactulose, polyethylene glycol, erythritol, sorbitol, xylitol): All act by drawing water into the bowel. Consequence is additive osmotic diarrhoea, which can be marked when both are taken at full dose; avoiding trehalose on laxative days limits it.

  • Berberine and fermentable fibre supplements (inulin, fructo-oligosaccharides): Inferred from berberine’s shared AMPK activation and loose stools and from fibre’s additive colonic fermentation. Consequence is additive bloating, flatulence and diarrhoea; introducing each separately at a low dose limits it.

  • Broad-spectrum antibiotics (clindamycin, ciprofloxacin, ceftriaxone): Antibiotic disruption of gut bacteria is the precondition for Clostridioides difficile expansion in the mouse models. Deferring trehalose until the microbiome recovers removes the inferred exposure.

  • Fasting, fasting-mimicking (very-low-calorie diets that imitate fasting) and ketogenic (very-low-carbohydrate) protocols: Trehalose is split into glucose, supplying about 4 kcal/g. Consequence is a broken fast and suppressed ketone production, undermining the autophagy these protocols seek; confining trehalose to eating windows avoids it.

Populations who should avoid Trehalose:

  • Documented trehalase deficiency or trehalose intolerance. Rests on a case report of mushroom intolerance resolving with trehalose avoidance (Bergoz 1971); biopsy series at activity around 6 IU/g protein record the deficiency without symptoms (Gudmand-Høyer 1988).

  • Recent Clostridioides difficile infection, within 90 days, or recurrent infection. Inferred from the mouse virulence model (Collins 2018) and the antibiotic-disruption precondition; the human cohort data found no association (Eyre 2019), and there is no human evidence of harm.

  • Active diarrhoeal illness, short bowel (surgically shortened intestine), or small intestinal bacterial overgrowth. Inferred by analogy from a review showing sorbitol absorption depends on absorptive surface and overgrowth ferments sugars into the same symptoms (Montalto 2013); no human data on trehalose in these groups.

  • Intravenous trehalose outside a clinical trial. Rests on regulatory status: parenteral (injected) trehalose, SLS-005, is an investigational product (NCT05136885), and the only large randomized trial showed an excess of serious adverse events (HEALEY 2025).

Risk Mitigation Strategies

Doses and timings below follow common practice unless cited.

  • Low starting dose: Starting at 3.3 g/day, the dose of the low-dose randomized trials (Yoshizane 2020), rather than the 100 g/day that produced withdrawals, avoids the dose-dependent bloating and diarrhoea.

  • Slow titration with a single-dose cap: Increases of 3 g every one to two weeks, with any single dose kept at or below 25 g, the single dose given in the glycaemic crossover trial (Yoshizane 2017), prevent osmotic diarrhoea.

  • Dosing with meals: Dividing the daily amount across meals slows delivery to the small intestine and was the condition under which the glucose and insulin attenuation was measured (van Can 2012). Reduces both osmotic load and glucose spiking.

  • Stopping on loose stools, re-challenging lower: Persistent diarrhoea at any dose signals that intake exceeds trehalase capacity. Stopping resolves it; resuming at half the dose avoids the dehydration and electrolyte loss of continued diarrhoea.

  • Genotyping before escalation: TREH rs2276064 genotyping identifies the reduced-activity allele before a dose escalation (Kozlov 2023); breath tests are unvalidated (Montalto 2013). Avoids attributing malabsorption symptoms to another cause.

  • Counting the calories: Subtracting trehalose calories, about 4 kcal/g, from the day’s intake prevents the weight gain that erased the vascular effect in the 100 g/day vascular trial (Kaplon 2016).

  • Deferral around antibiotics: Pausing trehalose during and for two to four weeks after a course of broad-spectrum antibiotics removes the inferred window of Clostridioides difficile susceptibility.

  • No parenteral use outside a trial: Parenteral trehalose remains investigational (NCT05136885); confining injected trehalose to trials avoids the serious-adverse-event profile of high-dose intravenous administration (HEALEY 2025).

Therapeutic Protocol

Parameters other than the cited doses — timing, titration steps and cycling — reflect common practice rather than tested regimens.

  • Low metabolic dose: 3.3 g/day, the dose of the glucose-tolerance trial in healthy volunteers and of the fracture-healing trial, taken with a meal for 12 weeks (Yoshizane 2020; Mehrvar 2026).

  • Mid metabolic dose: 10 g/day with meals for 12 weeks, the dose whose glucose-tolerance gain was measured only against baseline, in adults with a body mass index of 23 or above (Mizote 2016), and of the current microvascular trial (NCT05593549).

  • Substitution dose: 25 g as a single serving replacing an equivalent amount of glucose, the comparison tested for glycaemic and hormone response (Yoshizane 2017).

  • High autophagy-targeting dose: 100 g/day, the only dose tested for a vascular ageing endpoint, which produced withdrawals for gastrointestinal effects and a null primary comparison (Kaplon 2016).

  • Exercise dose: 30 g per hour as a drink during endurance exercise, the regimen that improved time-trial work after 100 minutes of cycling (de Oliveira 2024); the Osaka trials gave 500 mL of a 4–8% solution mid-exercise (Hamada 2022).

  • Competing approaches — parenteral: Investigational regimens bypass the gut entirely: 15 g or 30 g of 10% solution weekly (Zaltzman 2020) and 0.75 g/kg weekly (HEALEY 2025). Neither is available outside a trial.

  • Competing approaches — topical: 3% ophthalmic solution, usually with 0.15% hyaluronic acid, dosed four to six times daily, is the route with the largest controlled evidence base, per a systematic review (Ballesteros-Sánchez 2023).

  • Who popularised each approach: The metabolic doses come from Hayashibara’s own programme; the high-dose vascular approach from the Seals laboratory at Colorado; the parenteral approach from Bioblast Pharma and later Seelos Therapeutics.

  • Best time of day: No trial compared timings. Trials dosed with meals; splitting across meals and avoiding a large dose close to bedtime is common practice to limit overnight osmotic symptoms.

  • Half-life: Not established in humans. Oral trehalose is split in the intestinal lining within the absorption window, so, per a review, systemic exposure to intact trehalose is brief and low (Chen 2023).

  • Single versus split dosing: Split dosing is standard practice at any total above about 10 g/day, because tolerance is governed by the amount of trehalose arriving at the intestinal lining at once rather than by the daily total.

  • Genetic polymorphisms: TREH rs2276064 reduces trehalase activity and should lower the tolerated dose (Kozlov 2023). No pharmacogenetic dosing study exists for trehalose.

  • Sex-based differences: None established. No trial reports dose-response by sex, and the exercise trials enrolled men only, so the same doses are used for both sexes in practice.

  • Age-related considerations: The only trial in adults over 50 used the highest dose and lost participants to side effects (Kaplon 2016); starting at the low dose and titrating more slowly is the usual adjustment for older adults.

  • Baseline biomarkers: A between-group difference appeared only in participants whose two-hour glucose sat at the upper end of normal; the full sample showed none (Yoshizane 2020). Measuring it first tells whether there is anything to move.

  • Pre-existing conditions: Established type 2 diabetes showed no glycaemic change over 12 weeks at 3.3 g/day (Hashemian 2023); bowel disease lowers the tolerated dose independently of the target.

Discontinuation & Cycling

  • Lifelong versus short-term: Not resolved. The longest trials ran six to twelve months (Yap 2024; Zaltzman 2020); no lifelong regimen has been tested, and use is framed as a dietary substitution.

  • Withdrawal effects: None reported. No trial recorded a rebound or withdrawal syndrome, and the four-week washout in the 10 g/day trial saw parameters drift back toward baseline without symptoms (Mizote 2016).

  • Tapering: Not applicable for the metabolic effects. Gastrointestinal symptoms are dose-related and resolve on stopping, so a taper serves only comfort, not safety.

  • Cycling for efficacy: Not studied. No trial compared continuous with intermittent dosing, and no tolerance or receptor-downregulation mechanism has been proposed that cycling would address.

  • What reverses on stopping: The glycaemic effect is a same-meal substitution effect and ends with the last substituted dose; the 12-week glucose-tolerance change was not followed long enough after washout to say whether it persists.

Sourcing and Quality

  • Form to look for: Food-grade α,α-trehalose dihydrate powder, typically 98% or higher purity. The enzymatic starch-derived material from Hayashibara is the source behind most bulk resale and behind GRAS Notice GRN 45, which the FDA closed without questions.

  • Third-party testing: Trehalose is sold as a food ingredient rather than a tested supplement category, and no ConsumerLab product review of trehalose exists. A certificate of analysis showing identity, purity and heavy-metal and microbial limits is the practical substitute.

  • Particulate impurities matter for injectables: Nanoparticulate impurities in trehalose sold as a pharmaceutical excipient (inactive formulation ingredient) provoked an early immune response in laboratory models (Wang 2023), which is one reason food-grade powder is not a substitute for a parenteral product.

  • Parenteral and ophthalmic products: No compounding pharmacy supplies an approved intravenous trehalose; the investigational material was supplied by Seelos Therapeutics. Ophthalmic 3% trehalose is sold as finished eye drops, not compounded.

  • What to avoid: Products where trehalose appears only as an excipient or bulking agent in a blend, since the label amount is usually a fraction of a gram and far below any dose that has been tested.

Practical Considerations

  • Time to effect: The glycaemic substitution effect is immediate, within the same meal. The glucose-tolerance, inflammation and bone-healing changes were measured at 12 weeks, so that is the realistic trial length for anything beyond a single meal.

  • Common pitfall — assuming oral equals injected: The autophagy results come from cells, injected animals and intravenous trials. Oral trehalose failed to raise blood trehalose or deliver metabolic benefit even in trehalase-knockout mice (Yeh 2025).

  • Common pitfall — treating it as calorie-free: Trehalose is about 45% as sweet as sucrose but carries the same roughly four calories per gram, so it is a sugar swap, not a sweetener swap.

  • Common pitfall — starting high: The 100 g/day regimen is the one most often cited for longevity effects and is also the one that produced withdrawals and weight gain, with a null primary result (Kaplon 2016).

  • Regulatory status: A food ingredient, not a drug. The FDA closed GRAS Notice GRN 45 in 2000 with no questions and the European Union authorised it as a novel food in Commission Decision 2001/721/EC. No approved therapeutic indication exists.

  • Cost and access: Neither exceptionally expensive nor hard to obtain. Bulk food-grade powder is widely sold; the constraint is on the parenteral form, which is unavailable outside a trial.

Interaction with Foundational Habits

  • Sleep: No direct interaction and no human data. The indirect route is osmotic: an undigested fraction arriving in the colon causes overnight bloating and urgency that fragments sleep. Keeping the last dose several hours before bed and splitting large amounts across the day is the practical measure.

  • Nutrition: Direct and the main point of use. Replacing glucose with trehalose flattens the glucose and insulin curve of the same meal (van Can 2012), but the carbohydrate and calories still count. Taking it with food rather than alone also improves tolerance.

  • Exercise: Potentiating during endurance work. Taken as 30 g per hour during prolonged cycling it improved subsequent time-trial output over water, with flatter blood glucose than maltodextrin (de Oliveira 2024), and a second group replicated the late-stage effect (Hamada 2021). No advantage over other carbohydrates.

  • Stress management: Indirect and unconfirmed. In type 2 diabetes, stress-questionnaire scores were lower on 3.3 g/day than on placebo, but the between-group change was not significant (Hashemian 2023). A review of animal work ties trehalose to Nrf2 (a protein that switches on the cell’s antioxidant defences) (Pupyshev 2022); cortisol is unmeasured.

Monitoring Protocol & Defining Success

Before starting, the useful baseline is a two-hour plasma glucose on a 75 g oral glucose load, where a glucose-tolerance effect is the reason for use, since the response was confined to people whose two-hour value sat at the upper end of normal (Yoshizane 2020). Alanine aminotransferase gives a liver-safety baseline, because this enzyme shifted in an intravenous trial (Jamialahmadi 2022). High-sensitivity C-reactive protein is worth capturing where inflammation is the target. Body weight belongs in the baseline set too: weight gain erased the one vascular benefit measured (Kaplon 2016), so it checks whether the added calories offset the intended effect. Ongoing, the practical cadence is weight and digestive tolerance weekly during titration, then the full panel at 12 weeks — the duration of every trial that reported a change — and every 6 to 12 months thereafter if use continues.

Biomarker Optimal Functional Range Why Measure It? Context/Notes
Two-hour plasma glucose, 75 g oral glucose tolerance test No established target — track the change from the individual’s own baseline Marker the intervention is expected to change; the glucose-tolerance endpoint of the 12-week metabolic trials Diagnostic cutoffs exist but are not a reference range. Morning test after an 8-hour fast; the response was confined to people starting at the upper end of normal
Alanine aminotransferase 7–56 U/L (standard reference range) Safety check — a sustained rise from baseline stops use Alanine aminotransferase is a liver enzyme that rises when liver cells are injured. Avoid testing within 48 hours of hard exercise
High-sensitivity C-reactive protein No established target — track the change from the individual’s own baseline Marker the intervention is expected to change Published cut-points are risk categories, not a reference range. Invalid during any acute infection; repeat two weeks after recovery

Qualitative markers worth tracking:

  • Digestive comfort — bloating, flatulence and abdominal cramping, which are the first and most reliable signal that the dose exceeds intestinal enzyme capacity
  • Stool form and frequency, since loose stools mark the tolerance ceiling more sensitively than symptoms alone
  • Perceived energy levels, which changed in both arms of the vascular trial (Kaplon 2016) and so should not be read as a treatment effect
  • Eye comfort and dryness, where topical trehalose is being used

Emerging Research

  • Autophagy and microvascular function in type 2 diabetes: NCT05593549, early phase 1, 60 participants, 10 g/day for 14 days, measuring nitric-oxide-mediated vasodilation and autophagic flux in blood cells; recruiting to 2029. A positive result would lift the endothelial benefit above Low; a null would leave it a single after-the-fact subset finding.

  • HEALEY ALS platform, trehalose regimen: NCT05136885, phase 2/3, 161 participants, completed August 2023 with results posted and published. The null result is already reflected in this review: it caps any claim that trehalose, by either route, slows neurodegeneration in people.

  • STRIDES in spinocerebellar ataxia: NCT05490563, phase 2b/3 of intravenous trehalose, terminated November 2023 with 23 enrolled, for business reasons rather than safety or efficacy, and no results posted. Completion would have been the first controlled test of intravenous trehalose on ataxia scores.

  • Trehalose as add-on therapy in bipolar depression: NCT02800161, phase 3, 60 participants, 70 g/day versus maltose; completed October 2022 with no results posted. Publication either way would be the largest high-dose oral dataset, and a null would weigh against systemic effects at that dose.

  • Trehalose in fatty liver disease: NCT03738358, 39 participants, 5 g/day for 12 weeks with liver fat on computed tomography; completed March 2018, no results posted. A positive result would carry the animal fatty-liver work into humans; a null would weaken it.

  • Trehalose versus glycine air-polishing in peri-implant mucositis (gum inflammation around a dental implant): NCT07140146, 40 participants, listed as recruiting but past its estimated June 2026 completion, with no registry update since February 2026. A positive result would extend the periodontal finding to implants; a null would confine trehalose powder to equivalence with existing powders.

  • Future research area — oral bioavailability: The decisive question is whether swallowed trehalose can reach cells intact. Yeh 2025 found that even removing trehalase did not make oral dosing work in mice, which if replicated would rule out most proposed oral mechanisms.

  • Future research area — trehalase-resistant analogues: Lactotrehalose reproduced the metabolic effects without expanding Clostridioides difficile in mice (Zhang 2020). These compounds, not trehalose itself, are where the patentable development effort now sits.

Conclusion

Trehalose is a cheap, widely available sugar with an unusual property: it survives conditions that destroy other sugars, and in cells and animals it appears to stir the housekeeping machinery that clears damaged proteins. That is why it is in eye drops, vaccines and frozen food, and why it attracted longevity interest.

In people the picture narrows sharply. The strongest replicated human findings are not about ageing: swapping trehalose for glucose flattens the blood sugar and insulin rise of a meal, and taken as a drink it improves output late in endurance exercise. Combination eye drops ease dry eye; trehalose-only drops rest on small trials. A single trial points to better bone healing after a hip fracture. The claims closest to the longevity rationale are the weakest: the arterial finding rests on a subgroup of one small trial, the cell-clearing effect has never been measured in a person, and the one large trial testing it in a protein-clumping nerve disease found nothing.

The main obstacle is physical. An enzyme in the gut wall splits trehalose before much of it gets anywhere, which is why developers built an injected form — and why that form, in the one large trial, came with more serious events than placebo. Digestive upset is the reliable effect of large doses.

Evidence quality is uneven and the incentives visible: several supportive trials come from the manufacturer, several eye-drop trials from the drop maker, and because trehalose cannot be patented, funding has moved to modified versions.

Top - Benefits - Risks - Protocol