How far did my sodium and potassium intake actually fall when my calories halved?

I would like to sanity-check the electrolyte question with numbers rather than by adding a supplement and seeing what happens. My hypothesis is that the reason people feel better on electrolyte sachets has less to do with magnesium than with the fact that their sodium intake has collapsed, and I would like to know whether the arithmetic supports it.

What I can supply: before starting, my intake was around 2,600 kcal of a fairly ordinary diet with a lot of restaurant food and ready meals. Now it is around 1,300 kcal, home-cooked, high protein, very little processed food. I have not measured sodium at either point because nothing I ate had a label on it in the second period.

Can someone estimate what my sodium and potassium intakes plausibly were and are, and what a defensible replacement would look like? I would rather understand the size of the gap than guess at a dose. I have normal blood pressure and no kidney or heart problems.

The arithmetic supports your hypothesis strongly. Your sodium intake has plausibly fallen by 65-80%, which is a much larger relative change than your calories, because you made two changes at once and they compound. Potassium has probably fallen less, and may even have risen in density. Here is the working.

Sodium: the two-factor collapse

Population survey data consistently attribute roughly 70-75% of dietary sodium in Western diets to processed food, restaurant food and food prepared outside the home, with something like 10-15% naturally present in food and 5-10% added at the table or in cooking. Typical Western intake is around 3,400-3,600 mg/day of sodium, equivalent to about 8.5-9 g of salt.

Take 3,400 mg/day as your before figure and decompose it:

• Processed and restaurant sources, 72%: 2,448 mg
• Naturally occurring, 13%: 442 mg
• Added in cooking and at table, 15%: 510 mg

Now apply the two changes.

Change 1: intake halved, 2,600 to 1,300 kcal. If nothing else changed, sodium would halve to about 1,700 mg.

Change 2: processed and restaurant food to near zero. That removes the 72% component almost entirely. What remains is the naturally occurring fraction plus whatever you add in cooking, both scaled to your new intake:

• Naturally occurring, scaled: 442 × (1,300 / 2,600) = 221 mg
• Added in cooking, if you salt food normally: 400-600 mg. Call it 500 mg.
• Estimated current intake: about 720 mg/day

So 3,400 → 720 mg, a fall of 2,680 mg, or 79%. In salt terms that is roughly 8.5 g of salt down to about 1.8 g. Even if you salt generously and my cooking estimate is double, you land near 1,200 mg, still a 65% reduction.

For context: physiological requirement is low, in the region of 500 mg/day, so you are not deficient in the textbook sense. But requirement in a sedentary person at fluid balance is not the same as adequacy for someone who sweats, exercises, has had episodes of GI loss and has a reduced extracellular volume from weight loss. Sodium is the principal extracellular osmole; total body sodium is what holds plasma volume up. That is why the symptoms of low intake are postural light-headedness, poor exercise tolerance and headache rather than anything you would call a deficiency.

Potassium: the opposite direction

Potassium comes overwhelmingly from unprocessed food: vegetables, fruit, dairy, meat, potatoes, legumes. Typical intake is around 2,500-3,000 mg/day against a reference intake commonly given as 3,500 mg, so most people are below target to begin with.

• Before, at 2,600 kcal of processed-heavy food: perhaps 2,600 mg, i.e. 1,000 mg per 1,000 kcal.
• Now, at 1,300 kcal of home-cooked high-protein food, the density is likely higher, perhaps 1,300-1,500 mg per 1,000 kcal, giving 1,700-1,950 mg.

So potassium fell by roughly 25-35% in mass while improving in density. That is a real reduction but nothing like the sodium collapse, and it explains why the two behave differently. Meat is a decent potassium source, so a high-protein diet partly protects it.

What a defensible replacement looks like

Given normal blood pressure and no renal or cardiac disease, which is the condition on all of this:

• Sodium. The gap is roughly 1,500-2,500 mg/day of sodium relative to your previous intake. You do not need to restore all of it, and there is no reason to target the population mean, which is higher than anyone needs. Something in the range of an extra 1,000-1,500 mg/day, i.e. 2.5-4 g of salt, brings you into a range that is unremarkable by any standard. That is achievable by salting food, which is cheaper and more palatable than any sachet. Adding a further 300-500 mg around a training session in hot conditions is defensible.
• Potassium. Food, not supplements. The gap of a few hundred milligrams is closed by a portion of potatoes, a banana, some yoghurt or a larger serving of vegetables. Potassium supplementation is restricted in most jurisdictions for good reason and should not be self-prescribed at scale.
• Magnesium. Not addressed by this arithmetic because dietary magnesium is harder to estimate and serum levels are a poor guide. A time-limited trial at a modest dose is reasonable; indefinite use on the assumption of a deficiency you never measured is less so.

Why this matters for the original hypothesis

Your hypothesis was that electrolyte products work through sodium rather than magnesium, and the numbers are consistent with it. A typical electrolyte sachet contains several hundred to a thousand milligrams of sodium and a few tens of milligrams of magnesium. Given that your sodium gap is measured in thousands of milligrams and your magnesium status is unknown, the sodium is almost certainly doing most of the work. The characteristic tell is timing: sodium plus water improves postural symptoms within half an hour, which is a plasma-volume effect. A genuine micronutrient repletion does not act in thirty minutes.

The caveats are real, though. All of the above assumes normal blood pressure, normal renal function, no heart failure and no medication that alters sodium or potassium handling. Sodium restriction is prescribed for good reasons in a lot of people, and any of the conditions in that list changes the answer entirely. This is arithmetic on population averages applied to a stated set of circumstances, not advice.

edited 7 Jun 2026 by Dr_Tomas_Kral — added the method parameters

Worth adding what the sodium collapse does to fluid balance, because the two questions are usually treated separately and they are the same question.

Water follows sodium. Total body water is regulated to keep plasma sodium concentration in a narrow band, so if total body sodium falls, the body defends the concentration by excreting water, and extracellular volume falls. That is the mechanism by which a low-sodium diet reduces plasma volume, and it is also why the first few kilograms lost on any low-carbohydrate or low-processed-food diet are water: glycogen depletion contributes, but so does the sodium change.

The practical consequences that get misattributed:

• Drinking more water without any sodium can make postural symptoms worse, not better, because you dilute the remaining sodium and the kidney responds by excreting the water you just drank. People who are drinking three litres and still light-headed usually have a sodium problem, not a water problem. In extreme cases, high water intake with very low sodium intake and ongoing GI losses produces genuine hyponatraemia, which is dangerous.
• The scale reads a sodium change as progress. A 1-2 kg drop over a few days after cutting processed food is largely extracellular water. It comes back the moment you eat a restaurant meal, which people then interpret as having gained fat from one dinner.
• Exercise tolerance is plasma-volume sensitive. Stroke volume depends on venous return. A reduced plasma volume raises heart rate for any given workload, which is why people report their heart rate being higher for the same effort and conclude they have lost fitness. Some of that is real deconditioning; some is volume.
• Heat tolerance falls. Both because of plasma volume and because sweat sodium losses are harder to replace from a low-sodium diet.

The measurement that ties it together is the standing test described elsewhere here: heart rate lying and after two minutes standing. If the difference is large, the problem is volume, and volume means sodium and water together rather than either alone. It costs nothing and it is more informative than a serum sodium, which will usually be normal because concentration is what the body defends and volume is what it sacrifices.

A caution about the direction this line of reasoning is usually taken, since the arithmetic above is sound and its conclusion is easy to over-apply.

The finding is that sodium intake falls a great deal on this class and that modest repletion is reasonable in someone with normal blood pressure and normal renal function. The over-application is the idea that sodium is therefore unimportant to restrict and that everyone on a GLP-1 agonist should be adding salt. Three reasons that does not follow:

• The population overlap with hypertension is large. A substantial fraction of people taking these drugs have raised blood pressure, are on antihypertensive medication, or have heart failure or chronic kidney disease. In all of those groups sodium restriction is advised for reasons that have not been suspended. Someone on a diuretic and an ACE inhibitor adding salt sachets is not obviously doing the right thing.
• Blood pressure falls on this class anyway, substantially, through weight loss and probably through other mechanisms. That is one of the benefits. Deliberately raising sodium intake to treat the symptoms of a falling blood pressure is a strange trade if the falling blood pressure is the therapeutic effect. The better response to postural symptoms in someone on antihypertensives is usually a review of the antihypertensives, not more salt, and that is a clinician's call.
• The symptom is not specific. Everything in this topic presents as fatigue and light-headedness: volume depletion, energy deficit, anaemia, thyroid adaptation, over-treated blood pressure, poor sleep. Concluding sodium because sodium arithmetic is available is a form of looking under the streetlight.

The defensible version of the whole argument is narrow: if you have normal blood pressure, no renal or cardiac disease, no relevant medication, and postural symptoms that respond within half an hour to salt and fluid, then your sodium intake is a plausible cause and modest repletion is low-risk. Outside those conditions it is a conversation with someone who can see your blood pressure trend and your medication list.