The complete electrolyte handbook for fasting

Every electrolyte that matters during a fast — by name, with the mechanism, the dose, the supplement form, and an honest comparison of the commercial powders. The reference companion to our story-shaped electrolytes article.

There are seven electrolytes the human body cares about: sodium, potassium, magnesium, calcium, chloride, phosphate, and bicarbonate. Three of them — sodium, potassium, magnesium — do most of the work in a fasting context. Two more — phosphate and chloride — only become interesting under specific conditions. The last two — calcium and bicarbonate — usually take care of themselves, but it's worth knowing why.

This article walks each one in turn, then covers what to actually buy, what to mix yourself, when bloodwork matters, and which popular claims don't hold up.

What does fasting actually do to electrolytes?

A fast moves salt and water out of the body without ever being a story about either. The mechanism is the natriuresis of fasting: when you stop eating, insulin falls, and insulin's quieter job — telling the kidneys to hold sodium — falls with it. The distal nephron stops reabsorbing sodium efficiently, and the kidneys begin excreting it instead.

Three forces drive this together. First, insulin withdrawal directly reduces sodium reabsorption at the ENaC channel in the distal tubule (DeFronzo 1981). Second, ketoacid anions — β-hydroxybutyrate and acetoacetate — pull obligate cations like sodium and ammonium out with them; Sigler's classic 1975 J Clin Invest paper showed urinary sodium tracks ketoacid excretion at r ≈ 0.73. Third, the absence of filtered glucose in the proximal tubule reduces SGLT2-mediated sodium reabsorption (Tinawi 2020).

Cumulative sodium loss in the first few days of a true water fast can reach 200–400 mEq — the equivalent of 12–24 grams of salt. Renin and aldosterone are suppressed early in fasting, which makes the natriuresis stranger: the body is losing sodium despite the hormonal system that would normally retain it being switched off. The textbook "RAAS-driven" framing is wrong; the real mechanism is a coordinated three-way push at the tubule level.

For a 16-hour daily fast, none of this matters. For anything past 24 hours, sodium becomes the variable that determines whether you feel fine or feel awful.

How much sodium do you need during a fast?

Most multi-day fasters target 2–3 grams of sodium per day — about a teaspoon of table salt total, spread across the day. That range comes from Phinney and Volek's keto-adaptation work, which recommends 3–5 g/day for ketogenic adaptation, dialed down slightly for pure fasting where activity is usually lower (Volek & Phinney 2011).

This is well above the National Academy of Medicine's chronic-disease-risk-reduction target of 2.3 g/day, and it sits in active conflict with the AHA's blanket "less is better" framing. The conflict is unresolved in the primary literature: Mente's 2018 Lancet analysis of the PURE cohort found 3–5 g/day was associated with the lowest cardiovascular event rates in healthy adults, while Cook 2014 in Hypertension defended the lower target. Both sides agree the population guideline doesn't account for fasting-induced natriuresis specifically — that's a separate context the guideline never addressed.

The mechanistic case for higher sodium during a fast is straightforward. The kidneys are losing it. Replacing what you're losing keeps plasma volume and blood pressure stable. Symptoms of undershooting — frontal headache, lightheadedness on standing, heaviness in the legs, mild nausea — are almost always sodium plus water, not hunger.

There is, however, a credible counter-position. The TrueNorth medically-supervised water-fasting program runs 5–40 day fasts with no electrolyte supplementation at all, on the theory that the body autoregulates if no exogenous ions are introduced. A 2018 BMC Complementary and Alternative Medicine chart review of 768 patients reported low adverse-event rates under this protocol (Finnell 2018). The Goldhamer "no salt" position and the Phinney "load salt" position have never been compared head-to-head in a randomised trial. Both work for most people; both fail for some.

For self-directed fasts, 1–3 g of sodium per day is the safer middle ground — enough to blunt symptoms, low enough not to spike anyone's blood pressure unexpectedly.

Do you need to supplement potassium during a fast?

Potassium loss during fasting is real but small, and supplementing it is the most dangerous thing on this list. Most of the body's potassium sits inside cells; serum levels stay stable for a long time even with no intake, because the intracellular reservoir buffers the blood compartment.

Urinary potassium loss in early fasting is modest — driven by the same cation-anion balance with ketoacids that drives sodium loss, just at a much smaller magnitude. Symptomatic hypokalemia in healthy adults on water fasts of less than a week is rare. The risk inverts on refeed: insulin drives potassium back into cells, and serum potassium can crash post-meal. This is the danger window, not the fast itself.

The supplemental ceiling is unusually low. The FDA limits over-the-counter potassium chloride tablets to 99 mg per dose — about 2.5 mEq — because of small-bowel ulceration and arrhythmia risk at higher single doses. The Adequate Intake from food is 3,400 mg/day for men and 2,600 mg for women, but those numbers presume distribution across whole-food sources, not bolus supplementation.

For fasting, that means: 200–500 mg of potassium per day in a DIY mix is generally safe in healthy kidneys; 1–2 g/day from products like snake juice is in the contested zone; and supplemental potassium is contraindicated if you take ACE inhibitors, ARBs, potassium-sparing diuretics, or have any degree of chronic kidney disease. Cardiac arrhythmia risk emerges at serum K above 6.0 mEq/L.

The food-based alternative is safer: vegetables, broth, or even a small amount of low-sodium salt blend (which contains potassium chloride). Bone broth covers potassium and sodium together, with the bonus of being warm and savoury during a long fast.

Why magnesium often matters more than the others

Magnesium is unusual among fasting electrolytes because the deficit usually pre-dates the fast. Roughly half of US adults are below the Estimated Average Requirement at baseline (NHANES, NAM 1997). Modern soils and modern diets aren't generous with it. When fasting symptoms include muscle twitching, calf cramps at night, palpitations, or disrupted sleep, the fix is often correcting a chronic deficit — not addressing fasting-induced loss.

The kidney conserves magnesium efficiently during a pure water fast. Renal wasting is modest. The problem is that you stop taking any in, and a body that was already running below the RDA of 400–420 mg/day for men and 310–320 mg/day for women starts to feel it after 48–72 hours.

Magnesium has a separate clinical signature worth knowing: refractory hypokalemia. If serum potassium isn't responding to oral or IV potassium replacement, the missing variable is almost always magnesium — the renal tubule can't retain potassium without adequate magnesium. This is a textbook teaching point in nephrology, but it gets missed often enough that it's worth flagging here.

The supplement form matters more for magnesium than for any other electrolyte. Glycinate (bisglycinate) is well-absorbed, calming because glycine is an inhibitory neurotransmitter, and the standard recommendation for evening dosing. Citrate is well-absorbed, mildly laxative, and cheap — the right baseline form. Malate has the highest plasma AUC in some pharmacokinetic studies (Uysal 2019). L-threonate crosses the blood-brain barrier and is the only form with controlled-trial evidence for cognitive endpoints (Slutsky 2010 Neuron); it's expensive and contains relatively little elemental magnesium per dose. Oxide is the cheapest form and the worst — about 4% absorbed, primarily a laxative — and the supplemental UL of 350 mg/day is set specifically to avoid the diarrhea oxide reliably produces.

For fasting, 200–400 mg of glycinate or citrate before bed is the conventional protocol. Past 400 mg in a single dose, gastrointestinal side effects start to show up regardless of form.

What about calcium, chloride, phosphate, and bicarbonate?

Three of these four can usually be ignored during a daily-IF or short extended fast. The fourth — phosphate — has its own section below.

Calcium stays in normal range during fasts of weeks because parathyroid hormone and calcitriol defend serum calcium aggressively. The reservoir is enormous; bone holds 99% of body calcium and releases it when needed. Acute hypocalcemia on a fast is essentially never seen in healthy adults. The chronic concern is that prolonged fasting increases urinary calcium loss and bone resorption — relevant for repeat 5-day-plus fasters, irrelevant for someone doing 16:8. If supplementation is wanted, citrate is better-absorbed than carbonate and doesn't need gastric acid (Sakhaee 2001).

Chloride tracks sodium losses but at lower magnitude. Sigler 1975 specifically noted chloride loss is "much less than sodium" during fasting natriuresis. The clinical syndrome of hypochloremic metabolic alkalosis appears with vomiting, NG-tube suction, or prolonged fasting accompanied by vomiting — not with a normal fast. Salt provides chloride at a 1:1 molar ratio with sodium; the AI of 2.3 g/day mirrors the sodium AI by design. There is no reason to supplement chloride separately.

Bicarbonate is the body's primary plasma buffer, regulated by the kidney via reabsorption at the proximal tubule and de novo generation through ammoniagenesis. During fasting, mild metabolic acidosis from ketoacid accumulation pulls bicarbonate down; the kidney compensates by upregulating ammonium excretion, freeing bicarbonate to buffer the blood. Supplementing bicarbonate (the rationale for the baking soda in "snake juice" recipes) is mostly cargo-cult. Owen 1983 documented a transient post-refeed alkalosis that suggests endogenous bicarbonate generation actually overshoots. There's no controlled evidence that supplemental bicarbonate during a fast does anything except potentially blunt very early ketosis discomfort.

Refeeding syndrome — the danger window

Refeeding syndrome is the sudden electrolyte derangement that can occur when nutrition is reintroduced after a period of starvation. The defining mineral is phosphate. When carbohydrate hits an insulin-suppressed system, the insulin surge drives phosphate (along with potassium and magnesium) intracellularly to support ATP synthesis and 2,3-diphosphoglycerate production. Serum phosphate can crash within hours. Below 0.50 mmol/L (1.5 mg/dL), the syndrome produces cardiac failure, respiratory failure, rhabdomyolysis, arrhythmia, and seizure (Mehanna 2008 BMJ).

The risk threshold for healthy adults starts around the 5–7 day mark of a true fast. NICE guidelines flag five days of "little or no nutritional intake" as moderate risk, ten days as high risk. ASPEN 2020 stratifies by additional inputs — BMI, weight loss history, prior electrolyte abnormalities (Silva 2020). The hospitalised population for whom these guidelines were written is malnourished, anorexic, or chronically alcoholic — not someone on day five of a healthy water fast — but the underlying physiology is the same; the threshold is just lower for the at-risk groups.

The refeed protocol where risk applies is well-established: start at no more than 10 kcal/kg/day (5 kcal/kg in extreme cases), advance over 4–7 days, give 200–300 mg of thiamine before or with the first feed (Wernicke prophylaxis), and replace phosphate at 0.3–0.6 mmol/kg/day if levels drop. Monitor potassium, magnesium, phosphate, and calcium daily for the first week.

For typical IF and even 24–48 hour fasts in well-nourished adults, refeeding syndrome essentially does not occur — the literature isn't built on this population. For multi-day water fasters, the article on breaking a long fast covers the practical refeed protocol in more depth.

Which form should you take?

The supplement form matters for some electrolytes more than others. A short reference:

Mineral	Best form	Why	Avoid
Sodium	Any NaCl — table, sea, Himalayan	All ~39% Na by mass; trace minerals in unrefined salts are negligible	Sodium-supplement marketing claims about "60+ trace minerals"
Potassium	Food (vegetables, broth) or low-sodium salt blend (KCl)	Pharmaceutical KCl tablets are appropriate only with prescriber oversight	OTC KCl above 99 mg/dose; potassium gluconate megadoses
Magnesium	Glycinate or citrate, 200–400 mg/day	Glycinate is calming; citrate is cheap and well-absorbed	Oxide (~4% absorbed); sulfate (laxative)
Calcium	Citrate (if supplementing at all)	Absorbed independent of gastric acid; no meal required	Carbonate without food; dose >500 mg in one sitting
Chloride	Comes free with sodium chloride	No reason to supplement separately	—
Phosphate	Food on refeed; not supplemented during fast	Replacement is clinical, not OTC	—
Bicarbonate	Not supplemented for fasting	The body generates its own via the kidney	Sodium bicarbonate as a "fasting protocol" component

What to look for in an electrolyte powder

For a daily IF or 24-hour fast, none of this matters — a pinch of salt in water at zero cost does the job. For multi-day fasts, a pre-formulated powder is convenience, not necessity. Whatever's on your shelf works if it meets four criteria:

• Zero or near-zero calories. Anything past a handful ends the fast metabolically.
• No sugar. This is the hard line; see the next paragraph for why.
• Sodium ≥500 mg per serving for multi-day fasts, ≥250 mg is fine for shorter ones.
• Potassium and magnesium present at any non-trivial dose — exact amounts vary across products and don't matter much within the 100–500 mg range.

Many electrolyte powders on the market are not designed for fasting at all — they're designed for rehydration after exercise or illness. The two contexts look similar from the outside and are physiologically different. Rehydration products contain glucose on purpose: sodium-glucose cotransport via the SGLT1 transporter in the small intestine accelerates water absorption from the gut (Hunt 1992 Gastroenterology). That mechanism is the basis for the WHO Oral Rehydration Solution standard and saves lives in cholera-driven dehydration. It is useless for fasting, where the glucose itself ends the fast.

If a label lists sucrose, dextrose, glucose, or "natural cane sugar" in the first three ingredients, the product is a rehydration drink, not a fasting electrolyte. The simpler alternative is a packet that lists sodium chloride, potassium chloride, and magnesium citrate — and not much else.

What about DIY recipes?

The same dosing logic applies whether the powder is bought or mixed at home. Three common recipes:

Phinney-style mix (per ~16 oz water): ½ teaspoon fine salt (1.2 g sodium), ⅛ teaspoon low-sodium salt blend for potassium chloride (200–350 mg potassium), 150–200 mg magnesium glycinate or citrate powder, optional citric acid and stevia for taste. Built around Phinney and Volek's 3–5 g/day sodium target for keto adaptation; one or two of these per day covers most multi-day fasts. Costs cents per serving from bulk ingredients.

Salt-and-lemon water (per 12–16 oz): A pinch of salt (250–500 mg sodium) plus the juice of half a lemon (50 mg potassium, trace magnesium, citrate as a mild alkalinizer). Minimalist; sufficient for sub-24-hour fasts; undershoots sodium for anything multi-day.

Snake juice (Cole Robinson, per 2 L): 1 teaspoon potassium chloride (2 g potassium), ½ teaspoon table salt (1.2 g sodium), 1 teaspoon baking soda (1.2 g sodium as bicarbonate), ½ teaspoon food-grade Epsom salt (250 mg magnesium). The 2 g of potassium per batch is well above the OTC supplemental limit and is contraindicated with kidney disease, ACE inhibitors, ARBs, or potassium-sparing diuretics. The baking soda's rationale is rationale-light. Robinson is not a clinician; the recipe propagated through YouTube, not peer review. It's not unsafe in a healthy adult with normal kidneys — but it's not supported by trial data either.

For most fasters, the Phinney-style mix above — or any pre-made packet that meets the four criteria in the previous section — is the conservative choice.

When does this need bloodwork or a doctor?

Self-directed fasting with electrolyte supplementation is reasonable for healthy adults up to roughly 72 hours. Beyond that, or for anyone in the populations listed below, the rules change.

Signs that warrant a basic metabolic panel during or after a long fast: persistent severe headache past day three, palpitations or syncope, severe weakness or muscle pain (rhabdomyolysis screen), edema or rapid weight regain on refeed (refeeding syndrome workup), or confusion / ataxia / ophthalmoplegia (Wernicke encephalopathy — a medical emergency requiring urgent thiamine).

Cramping that doesn't respond to oral electrolytes is the classic signature of magnesium deficiency, which masks itself as refractory hypokalemia. If the cramps don't resolve with potassium replacement, the missing variable is almost always magnesium.

Who should not improvise this?

Five populations should not run extended fasts without a clinician:

• Anyone on diuretics (furosemide, hydrochlorothiazide, torsemide). Loop diuretics are aggressive magnesium wasters; thiazides waste potassium. Adding fasting-induced natriuresis on top can produce serious electrolyte derangement.
• Anyone with hypertension, especially salt-sensitive HTN or primary aldosteronism. Sodium loading at 3–5 g/day during a fast can produce dangerous BP spikes when blood pressure starts to rebound on refeed.
• Anyone on SGLT2 inhibitors (empagliflozin, dapagliflozin, canagliflozin). These drugs already cause urinary glucose, sodium, and water loss; combining them with a fast is a documented trigger of euglycemic DKA. The FDA recommends holding SGLT2 inhibitors three to four days before any procedure or fast.
• Pregnant or breastfeeding individuals. Extended fasting is contraindicated; even Ramadan-pattern intermittent fasting has shown modest fetal effects in observational studies.
• Anyone with chronic kidney disease (eGFR < 60). Reduced ability to excrete potassium and acid means potassium-containing electrolyte mixes can produce hyperkalemia, and metabolic acidosis worsens. Even daily IF benefits from supervision in this group.

Anyone with a history of disordered eating sits in a separate category — extended fasting is a clinical red flag regardless of electrolyte management.

Five common myths

"You don't need salt if you don't eat." False. Fasting increases sodium loss via insulin-withdrawal natriuresis. The mechanism is well-established (Sigler 1975; DeFronzo 1981). The question is dose, not whether.

"Athletes' electrolyte products work for fasting." False. Athletes lose sodium via sweat under glycogen-replete, insulin-normal conditions; sports drinks contain glucose because sodium-glucose cotransport accelerates water absorption from the gut. Fasters lose sodium via urine under insulin-suppressed conditions, and adding glucose ends the fast. Borrowing Gatorade-style numbers under-doses sodium and over-doses carbs.

"Pink Himalayan salt has meaningfully more minerals." False at culinary doses. Trace mineral content is sub-microgram per gram of salt. Same NaCl by mass as table salt. The "84 trace minerals" marketing is technically true and practically meaningless.

"You need bicarbonate to buffer fasting acidosis." Weak. Mild ketoacidosis is self-limiting in healthy fasting; the kidney generates its own bicarbonate via ammoniagenesis. Owen 1983 documented a post-refeed alkalosis that argues against routine supplementation. The baking-soda inclusion in snake juice is rationale-driven, not evidence-driven.

"Electrolyte powders are marketing hype." Half right. The premise — sodium replacement during fasting — is supported by mechanism. The specific brand differences are mostly marketing. Buy on price per gram of sodium; ignore the rest.

Sources

1. Sigler MH. "The mechanism of the natriuresis of fasting." J Clin Invest, 1975. PMC301756
2. DeFronzo RA. "The effect of insulin on renal sodium metabolism." Diabetologia, 1981. PubMed 7028550
3. Tinawi M. "Pathophysiology, evaluation, and management of metabolic acidosis." Front Endocrinol, 2020. PMC7221140
4. Mehanna HM, Moledina J, Travis J. "Refeeding syndrome: what it is, and how to prevent and treat it." BMJ, 2008. PMC2440847
5. Silva JSV, Seres DS, Sabino K, et al. "ASPEN consensus recommendations for refeeding syndrome." Nutr Clin Pract, 2020. PubMed 32115791
6. National Academies of Sciences, Engineering, and Medicine. "Dietary Reference Intakes for Sodium and Potassium." 2019. NBK538102
7. Institute of Medicine. "Dietary Reference Intakes for Calcium, Phosphorus, Magnesium, Vitamin D, and Fluoride." 1997. NBK109816
8. Sakhaee K, Bhuket T, Adams-Huet B, Rao DS. "Meta-analysis of calcium bioavailability: a comparison of calcium citrate with calcium carbonate." Am J Ther, 1999. PubMed 11329115
9. Uysal N, Kizildag S, Yuce Z, et al. "Timeline (bioavailability) of magnesium compounds in hours." Biol Trace Elem Res, 2019. PubMed 29679349
10. Finnell JS, Saul BC, Goldhamer AC, Myers TR. "Is fasting safe? A chart review of adverse events during medically supervised water-only fasting." BMC Complement Altern Med, 2018. PMC5819235
11. Owen OE, Caprio S, Reichard GA Jr, et al. "Ketosis of starvation: a revisit and new perspectives." Clin Endocrinol Metab, 1983. PMC292270
12. Slutsky I, Abumaria N, Wu LJ, et al. "Enhancement of learning and memory by elevating brain magnesium." Neuron, 2010. DOI:10.1016/j.neuron.2009.12.026
13. Mente A, O'Donnell M, Rangarajan S, et al. "Urinary sodium excretion, blood pressure, cardiovascular disease, and mortality." Lancet, 2018. PubMed 30152040
14. Volek JS, Phinney SD. The Art and Science of Low-Carbohydrate Living. Beyond Obesity LLC, 2011.
15. Related reading: Electrolytes — the quiet variable, How to break a long fast, and What breaks a fast — a decision tree.