No calories, no insulin response — the only universally neutral item.
No calories, no insulin response. The only universally neutral item under metabolic definitions. Still breaks most complete religious fasts (Ramadan, Yom Kippur, LDS).
Start here
Which Definition Applies
Find your reason for fasting below — it determines which rule actually governs your answer.
| Reason for fasting | What it measures | The operative rule |
|---|---|---|
| Religious / spiritual | Observance and intent, not physiology | Typically nothing by mouth at all, often including water. The tables below largely do not apply — see the religious section. |
| Fat loss | Caloric intake and insulin response | Items without meaningful calories that don't provoke insulin release fall outside the definition. Flavor alone doesn't count. |
| Autophagy / cellular repair | Amino acids, particularly leucine | Calories matter less here than protein content. Amino acids are the operative variable — though not all amino acids carry equal weight; see the label-reading guide below. |
| Blood sugar / insulin | Insulin response | Carbohydrate and protein both count. Fat is largely neutral. |
| Gut rest / digestive | Mechanical and digestive workload | Fiber, fat, and volume are the variables. Clear liquids are commonly permitted. |
| Medical — lab work | Whatever the ordering lab specifies | Follow the lab's written instructions. They supersede this page. Many specify water only. |
| Medical — pre-surgery | Aspiration risk under anesthesia | Follow the surgical team's instructions. Standard guidance permits clear liquids to 2 hours, solids 6–8. |
| “Clean fast” | Anything the body can detect | Water, black coffee, and plain tea comprise the permitted list. |
The distinction that resolves most online disagreement: a fat-loss definition asks “does it contain calories?” An autophagy definition asks “does it contain protein?” Those questions return different answers for the same item — which is why the same product gets opposite verdicts from equally informed people.
Key
Passes everywhereDoes not break a fast under any metabolic definition
Depends on definitionThe answer differs by lane — notes explain which
Breaks itBreaks a fast under all common metabolic definitions
Religious flagBreaks a religious fast even where metabolically neutral
The tool
Quick Check: Beverages, Foods & Additives
Search or filter to find an item, then tap it to expand the full reasoning.
Showing all 31 items
Carbonation is just CO₂ — unsweetened versions are metabolically equivalent to water.
Carbonation is dissolved CO₂. Unsweetened versions are metabolically equivalent to water. Some contain “natural flavors” — calorie-free, but excluded under strict clean-fast definitions.
Roughly 2–5 calories per cup. Caffeine modestly increases fat mobilization.
Roughly 2–5 calories per cup. Caffeine modestly increases fat mobilization and metabolic rate. Note the conflict: many labs require water only before bloodwork, though coffee is an approved pre-surgical clear liquid.
Calorie-free. Green tea catechins are linked to modest increases in fat oxidation.
Calorie-free. Green tea catechins are associated with modest increases in fat oxidation. Additions (milk, honey, sugar) change the answer.
Plain brewed mate is calorie-free. Canned mate is a different product.
Plain brewed mate is calorie-free. Canned mate is a different product — most contain sugar or juice.
~3 calories per tablespoon. May modestly blunt post-meal glucose response.
About 3 calories per tablespoon. May modestly blunt post-meal glucose response. Undiluted vinegar erodes tooth enamel and irritates the esophagus; dilution in water is standard practice.
No calories, no insulin response. Helps prevent fasting headache and fatigue.
No calories, no insulin response. Sodium loss is a recognized contributor to headache and fatigue during longer fasts.
Fat-loss: fine. Clean fast: breaks it — sweet taste is the excluded thing.
Fat-loss definition: does not break it — no calories, no meaningful insulin response in most people. Clean fast: breaks it; sweet taste is the specific thing excluded. Autophagy: no amino acids, so generally outside the definition.
Not interchangeable — LMNT is calorie-free, Liquid I.V. has ~11g sugar.
Not interchangeable products. LMNT: no calories, stevia-sweetened. Nuun: approximately 1g carbohydrate. Liquid I.V.: approximately 11g sugar, which breaks a fast under every metabolic definition. Unflavored electrolytes avoid the ambiguity.
No meaningful insulin response alone. Breaks a clean fast on taste alone.
Controlled trials show no meaningful insulin response from these compounds alone. Fat-loss definition: outside it. Clean fast: breaks it. A separate consideration is behavioral — sweet taste is associated with increased craving for some people. Some packets contain maltodextrin or dextrose as bulking agents, which are carbohydrates.
200–400 calories, minimal insulin response — but breaks autophagy and caloric fasts.
Roughly 200–400 calories with minimal insulin response. Ketogenic fat-loss definitions: often treated as permitted. Any caloric-restriction definition: breaks it. Autophagy: breaks it — calories suppress autophagy independent of protein.
20–60 calories, 2–8g sugar per serving. Breaks most definitions.
Typically 20–60 calories and 2–8g sugar per serving. Breaks a fast under most definitions; small volumes fall within some looser fat-loss definitions.
30–50 cal, 6–10g protein per cup — often assumed clear-liquid safe, but frequently isn't.
Roughly 30–50 calories and 6–10g protein per cup, which breaks autophagy and clean-fast definitions. It's often included in gut-rest protocols and fasting-mimicking approaches — but despite its reputation as a “safe” liquid, strained bone broth is protein- and fat-dense enough that many pre-surgical and medical clear-liquid protocols exclude it. Check the specific protocol rather than assuming broth passes.
45–60 calories, 9–12g sugar per cup — comparable to fruit juice.
45–60 calories and 9–12g sugar per cup. Nutritionally comparable to fruit juice.
Concentrated sugar without fiber. Rapid insulin response.
Concentrated fructose and glucose without fiber. Rapid insulin response. Pulp-free juice is an approved pre-surgical clear liquid — the sole lane where it passes.
Protein activates mTOR and suppresses autophagy, even in small amounts.
Protein content activates mTOR and suppresses autophagy. A tablespoon of cream is roughly 50 calories and 0.4g protein — enough to register.
7 cal/g. Ethanol metabolism halts fat oxidation; empty-stomach hypoglycemia risk.
7 calories per gram. Ethanol metabolism takes hepatic priority, halting fat oxidation. Alcohol on an empty stomach also carries hypoglycemia risk.
The most direct fast-breaker — amino acids terminate the fasted state.
The most direct fast-breaker. Amino acids are the specific signal that terminates the fasted state, regardless of calorie count. These are the established mTOR triggers (whey, collagen, BCAA/EAA blends) — see the label-reading guide below for which amino acids actually count.
~1 calorie per wedge. No measurable metabolic effect.
Roughly 1 calorie per wedge. No measurable metabolic effect. Lemonade and bottled juice are different products.
A dash carries negligible calories; cinnamon may modestly aid insulin sensitivity.
A dash carries negligible calories. Cinnamon is associated with modest improvements in insulin sensitivity.
Not swallowed, metabolically irrelevant — but breaks several religious fasts.
Not swallowed, metabolically irrelevant. Breaks a religious fast in several traditions — many observant Muslims avoid toothpaste during daylight hours in Ramadan.
Should never be stopped or delayed to preserve a fast.
Prescribed medication should not be stopped or delayed to preserve a fast. Timing is a question for the prescribing clinician. Some medications require food; some fasting protocols interact with insulin, sulfonylureas, and blood pressure medication.
Evidence is mixed — likely fine for fat loss, excluded before lab work out of caution.
Evidence is mixed. A controlled trial in fasted adults found no significant change in glucose or insulin after 30 minutes of chewing. A separate small clinical study (22 volunteers) reported shifts in cortisol, insulin, C-peptide, and triglycerides — a preliminary finding, not a settled one, but cautious enough that standard lab guidance excludes gum before a blood draw anyway. Practical reading: outside a fat-loss definition, excluded before lab work, breaks a clean fast.
1–5 calories per mint — below the fat-loss threshold, but breaks a clean fast.
Roughly 1–5 calories per mint — below the threshold that meaningfully affects fat loss or insulin. A standard Tic Tac is largely sugar but small enough to be labeled 0g. Breaks clean and religious fasts; unlikely to affect a fat-loss definition.
Trace protein, a few calories at most — immaterial to fat loss.
Trace protein, a few calories at most. Immaterial to fat loss. Tablets and vegetable capsules avoid the question entirely.
Minimal calories or insulin effect, but real digestive workload — breaks gut rest.
Minimal digestible calories and little insulin response, but meaningful digestive workload — which breaks a gut-rest definition while remaining roughly neutral for fat loss.
Low calorie, but still food. Chewing and digestion end the fasted state.
Low calorie, but food. Chewing, swallowing, and digestion end the fasted state under every definition.
Prepared with oil and salt — enough fat and fiber to initiate digestion.
Prepared with oil and salt. Small servings still carry fat and fiber sufficient to initiate digestion.
Refined starch. Rapid glucose and insulin response.
Refined starch. Rapid glucose and insulin response.
Natural sugar is still metabolically sugar; fiber slows, doesn't eliminate it.
Natural sugar is metabolically sugar. Fiber slows absorption without eliminating it.
Built on a glucose or cane-sugar base. Tablets deliver the same nutrients sugar-free.
Glucose syrup, corn syrup, or cane sugar base. Tablets and capsules deliver the same micronutrients without the sugar.
No items match your search. Try a different term or reset filters.
The one everyone gets wrong
Energy Drinks & Pre-Workouts
The usual claim is that energy drinks are permissible and pre-workouts aren't, because pre-workouts contain BCAAs. The category doesn't predict the contents — the label does.
C4 Original (powder)
Contains no BCAAs. Formula: caffeine, beta-alanine, creatine nitrate, L-arginine, L-citrulline, N-acetyl-L-tyrosine, B-vitamins, sucralose. Its beta-alanine and arginine content is not an autophagy concern — those are weak, non-proteinogenic triggers at best, not the leucine-driven signal that actually suppresses it.
C4 Energy + Aminos
From the same manufacturer — contains 5g BCAA/EAA, the
established trigger. Two products, one brand, opposite verdicts. The label is the only
reliable input.
What to read on a label
Established mTOR / autophagy triggers
Leucine, isoleucine, valine, BCAA/EAA blends, whey or collagen protein, glutamine. Leucine is the primary driver — these reliably activate mTOR and suppress autophagy.
Weak or theoretical — commonly overstated
Beta-alanine, taurine, citrulline, arginine, tyrosine. Amino acids, yes — but non-proteinogenic or lacking good evidence of suppressing autophagy at typical supplement doses. Don't treat these as decisive on their own.
Calories & carbohydrate → breaks fat loss
Sugar, dextrose, maltodextrin, glucose, cluster dextrin, sucrose, honey, juice concentrate.
Sweeteners & flavor → breaks a clean fast only
Sucralose, acesulfame-K, stevia. No calories, no meaningful insulin response — but excluded by definition from a clean fast.
By product type
| Product type | Fat loss | Autophagy | Clean fast |
|---|---|---|---|
| Zero-sugar energy drink | Passes0–15 calories | DependsUsually outside it — genuine BCAA/EAA blends are the exception; taurine alone is a weak, unconfirmed signal | Breaks itSweeteners and flavor exclude it |
| Stimulant-only pre-workout | PassesNegligible calories | DependsBeta-alanine/citrulline/arginine are weak, theoretical triggers — not the established kind | Breaks itFlavored and sweetened |
| Amino or EAA pre-workout | PassesOn calories alone | Breaks itDecisively — EAAs are the established, leucine-driven trigger | Breaks itFlavored, sweetened, and full of aminos |
| Carbohydrate-loaded pre-workout | Breaks itMeaningful sugar/dextrose load | Breaks itCalories and established aminos both present | Breaks itFails on every count |
| Plain caffeine tablet | PassesNo calories | PassesNo amino acids, no mTOR activation | PassesNo flavor, no sweetener |
Check a specific product against your goal
Pick a product above to see its verdict.
Terms decoded
An amino acid that buffers acid accumulation in muscle, extending endurance. It's responsible for the tingling, prickling sensation across the face and neck that arrives about fifteen minutes in — that sensation is paresthesia. It's harmless, temporary, and not an allergic reaction. It's non-proteinogenic (not used to build muscle protein), and there's no good evidence it suppresses autophagy at typical supplement doses — treat it as a weak, theoretical concern rather than a decisive one, unlike leucine or BCAAs.
Amino acids that raise nitric oxide and widen blood vessels, producing the “pump.” Same category as beta-alanine: technically amino acids, but weak or theoretical mTOR activators with no strong evidence of suppressing autophagy at supplement doses — not in the same tier as leucine or BCAAs.
Not a protein-building amino acid and not a meaningful insulin trigger. Plain unflavored creatine monohydrate is widely treated as neutral across metabolic definitions. Creatine nitrate and flavored blends carry other ingredients that may not be.
An amino acid, though not one used in protein synthesis, and — like beta-alanine, citrulline, and arginine — only a weak, theoretical mTOR activator. Genuinely ambiguous rather than a clear break; excluded under strict definitions out of caution.
The body's growth switch (mechanistic target of rapamycin). Amino acids activate it. Fasting deactivates it. Autophagy runs only while it's off.
Literally “self-eating.” Cells break down damaged components and recycle the materials. The process increases during nutrient scarcity, which is the principal reason people undertake longer fasts.
A different instrument entirely
Religious Fasting
Most of the guide above doesn't apply to religious observance. Metabolic reasoning is the wrong tool — the rules come from the tradition, and where a question is genuinely ambiguous, the answer comes from an imam, rabbi, priest, or bishop rather than a website. For orientation only:
Ramadan (Islam)
Nothing by mouth from dawn (Fajr) to sunset (Maghrib), including water. Coffee, gum, and mints all break it. Many observant Muslims also avoid toothpaste during daylight hours. Rulings on injections, inhalers, and other non-oral routes vary between schools of thought.
Yom Kippur (Judaism)
A complete fast from food and drink, approximately 25 hours.
Latter-day Saints
Typically 24 hours without food or water, commonly observed monthly.
Bahá'í Fast
No food or drink from sunrise to sunset, across 19 days.
Catholic
Two distinct practices. Fasting on Ash Wednesday and Good Friday permits one full meal plus two smaller ones. The Eucharistic fast is one hour before receiving Communion, excepting water and medicine.
Orthodox Christian Lent
Abstention from specific categories (meat, dairy, fish, oil, wine) rather than from all food — a structurally different practice from the above.
Every tradition listed here maintains established exemptions for illness, pregnancy,
nursing, travel, age, and medication requirements. Those exemptions are part of the
observance, not a departure from it.
Instructions override this page
Medical Fasting
Before bloodwork
The ordering lab's written instructions govern. They supersede everything on this page.
One conflict worth naming: black coffee is metabolically trivial and permitted under fat-loss definitions, but many labs require water only before lipid and metabolic panels, because coffee can shift triglyceride and glucose readings. Sugar-free gum is likewise excluded before a draw — a small study reported that it altered cortisol, insulin, C-peptide, triglycerides, and amylase, and labs tend to err cautious even though a single small study doesn't settle the question.
Before surgery or sedation
The surgical team's instructions govern. For context, standard anesthesia guidance for healthy adults undergoing elective procedures — per Joshi GP et al., the ASA's 2023 modular update to its Practice Guidelines for Preoperative Fasting (Anesthesiology 2023;138(2):132–151), which reaffirms the 2-hour clear-liquid window:
- Clear liquids — water, black coffee, black tea, pulp-free juice2 hours
- Light meal6 hours
- Fatty or heavy meal8 hours
- Non-human milk6 hours
Note how thoroughly this inverts fasting-community convention: pulp-free juice is approved, while adding milk to coffee moves it from the 2-hour category to the 6-hour one. The variable being managed is aspiration risk under anesthesia, not insulin.
Including where it's weaker than usually claimed
The Evidence
Tap a topic to see what's actually established versus what's a reasonable inference.
The frequently cited study is Pietrocola et al., “Coffee induces autophagy in vivo,” Cell Cycle, 2014, which found increased autophagic flux in liver, muscle, and heart tissue within 1–4 hours, alongside mTORC1 inhibition. Two qualifications rarely accompany the citation: the study was conducted in mice — human data remains limited — and decaffeinated coffee produced the same effect, meaning caffeine is not the active compound; the authors attribute the result to polyphenols. The supportable claim is that coffee appears not to interfere with a fast, with promising animal evidence suggesting benefit — not that it triggers autophagy in a given human on a given morning.
Well established — for the proteinogenic branched-chain amino acids. Leucine and other BCAAs activate mTORC1, which suppresses autophagy. The cell and animal biology is replicated and forms the mechanistic basis for the protein rule. The honest qualification: precise dose and duration thresholds in living humans are not established. “One gram of BCAA immediately terminates autophagy” is a reasonable inference rather than a measured human finding. The direction is certain; the threshold isn't. That certainty does not extend to every amino acid — beta-alanine, taurine, citrulline, arginine, and tyrosine are chemically amino acids but weak or unproven mTOR activators; see the label-reading guide in the Energy & Pre-Workout section.
Genuinely mixed, and smaller than it sounds. A controlled trial in fasted healthy men found no significant differences in glucose, insulin, or GIP following sugar-free gum. A separate small clinical study — 22 volunteers — reported shifts in cortisol, insulin, C-peptide, triglycerides, and amylase; it's the basis for standard guidance excluding gum before blood draws, but the sample size is too small to call the finding settled. The reasonable reading: gum may produce small, short-lived changes — enough that labs exclude it out of caution, not enough that it's established to affect fat loss.
The lab-draw caution and the surgical-fasting stance now point in different directions, which is a clean example of this page's whole thesis. Joshi GP et al.'s 2023 ASA preoperative fasting update advises against delaying surgery in otherwise healthy adults solely because they chewed gum beforehand — a more confident, more permissive conclusion than the lab-draw guidance above. Same substance, two medical contexts, two different risk calculations. Neither lane is wrong; they're just weighing different stakes.
The common claim that sweeteners cause insulin resistance is not supported at the confidence level it's usually stated. Established: sucralose, aspartame, and stevia produce no meaningful insulin or glucose response independently, including in people with type 2 diabetes. Contested: whether long-term use alters gut microbiota in ways affecting glucose tolerance. Research exists — notably Suez et al., Nature, 2014, and subsequent work — but findings are inconsistent and effects vary substantially between individuals. This is an open question rather than a settled finding.
The 2-hour clear-liquid window is formal clinical guidance most recently reaffirmed in Joshi GP et al., the ASA's 2023 modular update to its Practice Guidelines for Preoperative Fasting (Anesthesiology 2023;138(2):132–151), and endorsed in comparable form by the European Society of Anaesthesiology and Intensive Care (ESAIC). Standard guidance for healthy adults undergoing elective procedures: clear liquids (water, black coffee, black tea, pulp-free juice) — 2 hours; light meal — 6 hours; fatty or heavy meal — 8 hours; non-human milk — 6 hours. The variable being managed is aspiration risk under anesthesia, not insulin.
Scope
What This Guide Is — and Isn't
This page answers one question: whether a given item breaks a fast, and under which definition. It doesn't recommend fasting, prescribe durations, set intake targets, or provide protocols. Those are questions for a clinician who knows the individual's history.
Before fasting
Medical guidance is standard before fasting for anyone who:
- is pregnant or nursing
- is under 18
- is managing diabetes or taking glucose-lowering medication
- has a history of disordered eating
- takes prescription medication
- has kidney, liver, or heart conditions
If fasting has stopped functioning as a tool
…and started functioning as a rule that's hard to put down, that's worth talking to someone about. In the US, the National Alliance for Eating Disorders helpline is 1-866-662-1235.