Sodium is an essential nutrient that most Americans eat far more of than any guideline recommends, and almost none of the excess comes from the salt shaker. The controlled feeding trials are consistent: lower sodium, lower blood pressure. The observational literature is messier, and has produced a genuine argument about how low is too low. This article covers what the trials measured, where the sodium in a US diet originates, who responds most to cutting it, and what separates sea salt from table salt — which is not what this blog said in 2014.
How much Americans eat, and what the guidelines say
The FDA puts average US intake at about 3,400 mg of sodium a day. The Dietary Guidelines for Americans set a limit of less than 2,300 mg a day for adults, roughly one teaspoon of table salt. The American Heart Association sets an ideal of 1,500 mg for most adults.
Physiological need is far lower than any of these — a few hundred milligrams a day covers fluid balance, nerve conduction and muscle contraction. The gap between what is needed, what is recommended and what is eaten is the whole subject.
What the blood pressure trials found
The strongest evidence comes from controlled feeding, where investigators supply every meal rather than asking people what they ate.
The DASH-Sodium trial, published in the New England Journal of Medicine in 2001, randomized 412 adults to either a typical American control diet or the DASH diet, and within each, to high, intermediate and low sodium for 30 days each in random order. Reducing sodium lowered systolic blood pressure at both diet levels. Compared with the control diet at high sodium, the DASH diet at low sodium produced a mean systolic pressure 7.1 mm Hg lower in participants without hypertension and 11.5 mm Hg lower in those with it.
The CARDIA-SSBP trial, in JAMA in 2023, ran a crossover in 213 adults aged 50 to 75, a week on a high-sodium diet (about 2,200 mg added to usual intake) against a week on 500 mg total. Median systolic pressure was 126 mm Hg on high sodium and 119 on low. The difference at the end of the first intervention week was 8 mm Hg (95% CI 4 to 11). It fell by 5 mm Hg or more in 73.4 percent of participants, and the effect held whether or not they were taking blood pressure medication.
The trial that measured outcomes, not just pressure
Blood pressure is a surrogate. One large trial has randomized people to a sodium intervention and counted events.
The Salt Substitute and Stroke Study, published in 2021, was a cluster-randomized trial across 600 rural Chinese villages and 20,995 participants, all of whom had either survived a stroke or were 60 or older with high blood pressure. Villages received either regular salt or a substitute of 75 percent sodium chloride and 25 percent potassium chloride. Over roughly five years, stroke occurred at 29.14 events per 1,000 person-years on the substitute against 33.65 on regular salt — a rate ratio of 0.86 (95% CI 0.77 to 0.96). Major cardiovascular events and death from any cause were also lower.
The safety result matters as much. Serious adverse events attributed to high blood potassium ran at 3.35 per 1,000 person-years on the substitute and 3.30 on regular salt, a rate ratio of 1.04 (95% CI 0.80 to 1.37). This was a high-risk older population in one country, though, and potassium substitutes are not for everyone: anyone with reduced kidney function, or taking ACE inhibitors, ARBs or potassium-sparing diuretics, needs to ask a clinician first.
The J-curve, and why the argument is real
Several large observational studies have reported that both very high and very low sodium excretion track with worse outcomes — a J-shaped or U-shaped curve. The most cited is PURE, published in the New England Journal of Medicine in 2014, which measured 102,216 adults across 18 countries. Each additional gram of estimated sodium excretion was associated with 2.11 mm Hg higher systolic pressure, and the slope was much steeper at high intakes (2.58 mm Hg per gram above 5 g a day) than at low ones (0.74 below 3 g).
Two objections to reading a J-curve as causal are worth taking seriously. PURE estimated intake from a single fasting morning urine sample converted by formula, and single spot samples are an imprecise proxy for 24-hour excretion, with errors that flatten and distort the curve at both ends. And people who eat very little are often people who are already ill — reverse causation is difficult to exclude in a cohort.
Set against that, the Trials of Hypertension Prevention follow-up, published in the BMJ in 2007, tracked 3,126 adults who had been randomized to sodium reduction 10 to 15 years earlier and recorded 25 percent fewer cardiovascular events in the intervention group, or 30 percent after further adjustment. That is randomized allocation followed long enough to count outcomes, and it points the other way.
The honest position is that the trial evidence for moderate reduction is solid, the outcome evidence at very low intakes is thin in both directions, and anyone claiming certainty about the bottom of the curve is ahead of the data.
Where the sodium actually comes from
This is the part that changes behavior. More than 70 percent of dietary sodium in the United States comes from packaged and restaurant food, not from cooking or the table. The FDA names deli sandwiches, pizza, burritos, soups, savory snacks, poultry, pasta dishes, burgers and egg dishes as the categories accounting for roughly 40 percent of intake.
Bread is the quiet one: a slice does not taste salty, but Americans eat a lot of it. The practical consequence is that switching salts on the table adjusts a small fraction of a small fraction, while cooking at home changes the number that matters. Our overviews of diet and long-term disease risk and the Mediterranean diet cover the pattern this fits into.
Who is salt-sensitive
Blood pressure response to sodium varies widely between individuals. In CARDIA-SSBP a commonly used threshold classified about 46 percent of participants as salt-sensitive, while nearly three-quarters showed a systolic fall of at least 5 mm Hg — so the label depends heavily on where the line is drawn.
The American Heart Association’s 2016 scientific statement on salt sensitivity describes it as more common with older age, in Black adults, in people already hypertensive, and in those with chronic kidney disease, diabetes or obesity. There is no practical clinical test, which is why guidance is population-wide rather than personalized. The realistic approach for an individual is to cut intake for a few weeks and have blood pressure measured properly at both ends.
Sea salt versus table salt
The 2014 version of this article claimed unrefined sea salt contains over 80 essential nutrients and minerals, that it balances the body’s pH while table salt makes it acidic, and that readers should trade one for the other. All three claims were wrong, and we are withdrawing them. Blood pH is held in a narrow range by the lungs and kidneys, and is not adjustable by seasoning.
The trace minerals are real but nutritionally trivial: milligrams per hundred grams of salt, against daily requirements that would need an intake of salt no one could survive. A person eating sea salt for its magnesium is getting a rounding error.
| Table salt | Sea salt | Kosher salt | |
|---|---|---|---|
| Sodium by weight | About 40% | About 40% | About 40% |
| Sodium per teaspoon | Highest — fine crystals pack densely | Lower if coarse | Lower — flakes pack loosely |
| Iodine | Usually added | Usually none | Usually none |
| Trace minerals | Essentially none | Trace, nutritionally negligible | Essentially none |
| Microplastics | Lower | Higher in sampling studies | Depends on source |
Two differences do matter, and both favor ordinary table salt. The first is microplastic content. A 2018 study in Environmental Science & Technology analyzed 39 commercial food-grade salt brands, including 28 sea salts from 16 countries. Microplastic content ranged from 0 to 1,674 particles per kilogram in sea salt, with one outlier at 13,629, against 0 to 148 in rock salt and 28 to 462 in lake salt. Sea salt tracked measured seawater plastic pollution closely enough that the authors proposed it as an indicator of it.
The second is iodine, and it is the more consequential. US table salt has been iodized voluntarily since the 1920s and, per the NIH Office of Dietary Supplements, contains roughly 45 mcg of iodine per gram. Sea salt, kosher salt, Himalayan salt and fleur de sel are not usually iodized, and the salt used in processed foods almost never is. The adult RDA is 150 mcg a day, rising to 220 in pregnancy and 290 during lactation. Median urinary iodine in US pregnant women measured 144 mcg/L in 2007–2014 data, below the 150 mcg/L threshold for sufficiency in pregnancy. Swapping iodized salt for a specialty salt removes a nutrient with no substitute in most American diets, in exchange for minerals that are not doing anything.
Common questions
Is sea salt healthier than table salt?
No. Both are roughly 40 percent sodium by weight, the trace minerals in sea salt are nutritionally negligible, sampling studies have found more microplastic particles in sea salt than in rock salt, and sea salt usually lacks the iodine that iodized table salt supplies. If the texture and flavor are the point, that is a reasonable culinary preference. It is not a nutritional upgrade.
Does coarse salt contain less sodium?
Per teaspoon, yes; per gram, no. Coarse and flaked crystals pack less densely, so a teaspoon of kosher salt holds less sodium than a teaspoon of fine table salt. Weigh it rather than measuring by volume if the number matters.
Should I use a potassium salt substitute?
The SSaSS trial recorded lower stroke and mortality rates with a 75/25 sodium-potassium substitute in older, high-risk adults, without an excess of hyperkalemia events. But the study population was specific, and potassium loading is genuinely risky in reduced kidney function or alongside ACE inhibitors, ARBs and potassium-sparing diuretics. It is a question for a clinician, not a supermarket aisle.
Do athletes need more sodium?
Some do. Prolonged exercise in heat produces real sodium losses, and endurance events are where replacement earns its keep. For most people ordinary intake already exceeds anything a workout removes, as our piece on hydration and water quality covers.
What to do with all this
The evidence supports moving toward the guideline rather than chasing the last hundred milligrams: cook more at home, read labels on bread, soup and packaged meat, and eat enough potassium-rich plant food that the ratio improves from both ends. What it does not support is buying a more expensive salt and believing something has been fixed.
Vegetables and fruit are the potassium side of that equation, and most Americans fall short of them. Greens Plus has been making green superfood blends since 1993.
About the Author
