News · Nutrition & Diet
Who benefits from cutting salt varies 15-fold
Guidelines tell everyone to eat less sodium. Pooling 160 studies, researchers found blood pressure response depends heavily on age, starting pressure and body weight.
Based on a peer-reviewed metaregression of 160 studies in Hypertension
- Researchers pooled 160 studies with 255 separate estimates covering 16,443 people, publishing in the American Heart Association journal Hypertension.
- Across the whole population, each 50 mmol/day more sodium in the urine went with 1.76 mm Hg higher systolic pressure (95% CI, 1.54-1.98).
- Age and starting blood pressure were the strongest modifiers of that response.
- The effect grew by 0.41 mm Hg for every 5 years of age and 0.23 mm Hg for every 5 mm Hg of higher baseline pressure.
- A highly salt-sensitive profile was defined in advance as a systolic drop of at least 3 mm Hg per 50 mmol/day reduction.
- The number of people who need to cut sodium to prevent one cardiovascular event varied roughly 15-fold across groups.
- In older adults with high blood pressure and a raised body mass index, that number fell to about 87.
- This pools existing studies rather than testing individuals, so it predicts group averages, not whether any one person is salt-sensitive.
Every major guideline says the same thing about salt: everyone should eat less of it. A study in Hypertension asked a question that uniform advice tends to skip, which is how much the benefit actually differs between people, and found the answer spans more than an order of magnitude.
The authors set the problem out plainly. Hypertension guidelines generally recommend uniform sodium restriction although determinants of salt sensitivity, or sodium-blood pressure responsiveness, remain incompletely quantified.
Pooling six decades of sodium trials
Rather than run another trial, the researchers assembled the existing ones.
They searched PubMed, Embase, and Cochrane from inception to April 2025 for sodium intervention trials and observational studies, standardizing every result so it could be compared: blood pressure response was standardized per 50-mmol/d sodium difference, which is roughly 3 grams of salt.
The pool came to 160 studies (255 estimates; n=16,443).
They also set the threshold before looking. A highly salt-sensitive phenotype was defined a priori as a systolic blood pressure decrease of at least 3 mm Hg per 50-mmol/d reduction. Deciding in advance what counts as a large response is what stops a study like this finding whatever pattern it prefers.
The average effect
Across everyone, sodium mattered, and not by much per person.
Each 50-mmol/d higher urinary sodium excretion was associated with 1.76 mm Hg higher systolic blood pressure, with a range from 1.54 to 1.98.
Under two points. That is a genuinely small number for an individual and a genuinely large one for a country, because cardiovascular risk climbs continuously with blood pressure and shifting a whole population by two points prevents a lot of strokes.
The average conceals the story
The interesting part is how unevenly that 1.76 is distributed.
Age and baseline systolic blood pressure were the strongest modifiers, at 0.41 mm Hg per 5 years of age and 0.23 mm Hg per 5 mm Hg of starting pressure.
Stack those. A 70-year-old with high blood pressure sits many increments up both scales from a 40-year-old with normal pressure, and the responses compound rather than cancel.
The authors then translated that into something practical: the number of people who would need to cut their sodium to prevent one cardiovascular event. In older adults with hypertension and elevated body mass index - age 60 or over, systolic pressure 130 or higher, and body mass index 25 or above - that number fell to about 87, representing a 15-fold gradient across the strata examined.
Eighty-seven people is an efficient intervention by any standard. Fifteen times that is a different proposition entirely, even though the advice given is identical.
What this does not do
It pools studies rather than testing people. Metaregression works on group averages, so it can tell you that older people with higher pressure respond more on average; it cannot tell any individual whether they personally are salt-sensitive.
Salt sensitivity also has determinants this analysis could not capture, including genetics, kidney function and race, which have all been linked to sodium response. The three modifiers identified here are simply the ones routinely measured in the studies available.
Combining trials with observational studies brings its own difficulties, since the two designs are vulnerable to different biases and urinary sodium is itself an imperfect measure of what someone eats.
And nothing here shows that anyone’s response is zero. The population effect was positive throughout; what varied was how large.
Why it is useful
The conclusion is measured: salt sensitivity varied across age, systolic blood pressure, and body mass index strata, and these findings suggest that routinely measured clinical characteristics may help identify populations with greater expected blood pressure response to sodium reduction.
Populations, not individuals, and may help rather than will. What it changes is the framing of advice. Telling a healthy 35-year-old and a 70-year-old with hypertension the same thing about salt is not wrong, exactly. It is just that one of them is being handed a far better deal than the other, and nobody has been mentioning it.
People also ask
What does salt sensitivity mean?
How much a person's blood pressure moves when their sodium intake changes. It is a spectrum rather than a category. Some people can halve their salt and see almost nothing; others drop several points. The trait has been recognized for decades, but there has never been a practical way to test for it outside a research setting, which is part of why advice has been given uniformly.
So should people stop cutting salt?
No, and this is general information rather than medical advice. Sodium raised blood pressure across the whole population studied; the finding is that the size of the effect varies. Older adults, people with existing high blood pressure and people carrying extra weight had the most to gain. That is a large share of the people most at risk of heart disease in the first place, so the public health message is largely unchanged.
What does a 15-fold difference in number needed actually mean?
It is the number of people who would have to reduce their sodium to prevent one cardiovascular event. When that number is around 87, as it was in older adults with high blood pressure and raised body mass index, sodium reduction is a highly efficient intervention. When it is fifteen times larger, the same advice delivers far less per person following it. Same recommendation, very different value.
Is 1.76 mm Hg a meaningful change?
On its own it sounds trivial, and at a population level it is not. Shifting the entire distribution of blood pressure down by a couple of points prevents a substantial number of strokes and heart attacks, because risk rises continuously with pressure and the population is enormous. For an individual, a couple of points is unlikely to be noticeable.
Can someone find out if they are salt-sensitive?
Not easily. Formal testing involves controlled sodium loading and depletion under supervision, which is a research procedure rather than a clinical one. What this study offers instead is a way to estimate the likely response from characteristics a clinic already records: age, blood pressure and body mass index. That is a rough guide, not a test, and anyone with high blood pressure should discuss management with their clinician.