Claims review

Salt, the Second Pill, and the Ankles That Swell by Evening

Cutting salt is the first instruction almost everybody with a high reading is given. It works, and it works less than people are led to expect. Here is what the trials measured, why the top number keeps climbing with age anyway, and why the tablet that gets added next is a well-documented reason for ankles that swell by evening.

The order things usually happen in

She cut the salt, and the number did not move. It tends to go like this. A reading comes back high at a routine visit. The first instruction is salt: read the labels, stop adding it at the table, cook without it. So the salt cellar goes into the cupboard, the soy sauce gets measured, and the bread gets swapped for the low-sodium loaf nobody likes.

Three months later the number has come down a little, or not at all. A tablet is prescribed. A few months after that the number is still above target and a second tablet is added. And somewhere in the weeks that follow, the shoes that fit at breakfast are tight by six, and there is a ring around each ankle where the sock sat.

Most people in that position blame the one thing they were told to blame at the start, and cut the salt further. This page is about why that is usually the wrong suspect, and it goes through the sequence one step at a time, with what has been measured at each.

What cutting salt does, in numbers

Salt reduction is one of the better-studied things in all of nutrition, and the answer is not in dispute: it lowers blood pressure. The question is by how much.

The Cochrane review of longer-term, modest salt reduction pooled thirty-four trials and 3,230 participants. Eating about 4.4 grams less salt a day, for at least four weeks, lowered the top number by 4.18 and the bottom number by 2.06. In people who already had high readings the fall was larger, 5.39 and 2.82; in people with normal readings it was 2.42 and 1.00 — Cochrane Database of Systematic Reviews, 2013, DOI 10.1002/14651858.CD004937.pub2.

Two things are worth saying about those figures. The first is that 4.4 grams is a lot of salt to remove: it is most of a teaspoon, every day, and the trials measured it in urine and did not take anybody’s word for it. The second is that five points is an average. A reading that starts at 158 and falls by five is 153, which is still a long way above where a prescriber wants it. Somebody can do everything asked of them at the table, do it properly, and still be told the number is too high. That is not a failure of effort. It is the size of the effect.

For comparison, a change of the whole eating pattern does a little more. In the DASH trial, 459 adults were fed a diet rich in fruit, vegetables and low-fat dairy for eight weeks. It lowered the top number by about 5.5 points compared with a typical American diet, with salt held the same in both groups — New England Journal of Medicine, 1997, DOI 10.1056/NEJM199704173361601. Salt is one lever. It was never the only one.

The whole list, added up

Salt is rarely the only thing on the list. The note on the refrigerator usually has three lines: cut the salt, lose the weight, walk every morning. Each of the three has been measured, and each of them works.

Weight first. A meta-analysis of twenty-five randomised trials and 4,874 participants found that losing an average of 5.1 kilograms, about eleven pounds, lowered the top number by 4.44 and the bottom number by 3.57, which comes to roughly one point for each kilogram lost — Hypertension, 2003, DOI 10.1161/01.HYP.0000094221.86888.AE. Walking next. Across ninety-three trials and 5,223 participants, endurance exercise lowered the top number by 3.5 and the bottom by 2.5, and by 8.3 and 5.2 in the groups that started with hypertension — Journal of the American Heart Association, 2013, DOI 10.1161/JAHA.112.004473.

So a person who cut the salt, lost twenty pounds and walks every morning has done three things that each lower the reading, and nobody should take that away from them. What the trials do not promise is that the three results stack neatly on top of each other, or that the total is enough. Somebody who started at 168 and has earned back fifteen points by sheer effort is at 153. The effort was real. The target is 130. It is not only the salt and it is not only the weight, and being told to try harder at that point is an answer to a different question.

The neighbour who salts everything

Everybody knows one. Eighty years old, salt on the eggs, salt on the tomatoes, bacon most mornings, and a reading a forty-year-old would be pleased with. And across the street, somebody who has not touched the salt cellar in a year and whose number keeps climbing.

The two of them are not a contradiction, and they are not evidence that salt is harmless. Blood pressure responds to salt far more in some people than in others, and the difference is recognised well enough to have its own scientific statement — Salt Sensitivity of Blood Pressure, American Heart Association, Hypertension, 2016, DOI 10.1161/HYP.0000000000000047. The five-point figure above is an average across thousands of people. Inside that average are people whose reading falls a long way when the salt goes and people whose reading barely moves, and there is no test a household can do to find out which one it has.

That is the honest reason two people can eat at the same table and get opposite results. The difference between them is not on the plate. It is in how each body, and each set of arteries, handles what is on it.

Why the top number climbs with age anyway

The part that rarely gets explained at a ten-minute appointment is that the top number rises with age in most people whatever they eat, and the reason is mechanical.

The Framingham Heart Study followed 2,036 people for up to thirty years and plotted both numbers against age. The top number rose in a straight line from thirty to eighty-four. The bottom number rose with it until somewhere between fifty and sixty, and then turned round and fell, so that the gap between the two widened steeply. The authors’ reading of that pattern is in their conclusion: it is consistent with increased large artery stiffness — Circulation, 1997, DOI 10.1161/01.cir.96.1.308.

A young artery stretches a little with each beat and springs back between beats, which smooths the pressure out. An older one stretches less. The same volume of blood pushed into a stiffer tube produces a higher peak and a lower trough, and that is exactly the pattern on the chart: top number up, bottom number down. The same paper adds a sentence that is easy to miss and worth keeping, that a higher top number left alone may accelerate large artery stiffness and thus perpetuate a vicious cycle.

None of that is caused by a salt cellar, and none of it is undone by putting one in the cupboard. It is the main reason a person of sixty-five can eat carefully and still watch the reading drift upward year on year.

What the cuff cannot see

A blood pressure cuff does one thing. It squeezes the arm and reports how hard the blood is pushing against it. It is a measurement of force. It says nothing about the condition of the wall the blood is pushing against, and it was never built to.

Looking at the wall takes a different kind of test, and two of them are in ordinary use. An ultrasound of the arteries in the neck shows the thickness of the wall and any build-up on it. A CT scan of the heart can count the calcium in the coronary arteries and turn it into a score. Calcium is the part that surprises people, because the word they have heard at every appointment is cholesterol, and a cholesterol figure comes from a blood sample. Calcium in an artery wall is a mineral deposit, it shows up white on a scan, and no blood test reports it.

It matters because it predicts. In a study that scanned 6,722 adults with no known heart disease and followed them for a median of 3.8 years, people with a calcium score above 300 had 9.67 times the risk of a coronary event of people with no calcium at all, and the score added information beyond the usual risk factors — New England Journal of Medicine, 2008, DOI 10.1056/NEJMoa072100.

Two cautions belong next to that, and they are the difference between information and alarm. A calcium scan is not a routine test for everybody; whether it is worth having is a decision for a clinician who knows the rest of the picture, and it involves a small dose of radiation. And we know of no supplement that has been shown in a trial to remove calcium from an artery wall. The scan is a way of seeing more clearly. It is not a shopping list.

How common it is to be above target

Very, and the figures are public. The National Center for Health Statistics measured blood pressure in a national sample between August 2021 and August 2023. Hypertension, defined as 130 over 80 or above or taking medication for it, was found in 47.7 per cent of adults: 23.4 per cent of those aged eighteen to thirty-nine, 52.5 per cent of those forty to fifty-nine, and 71.6 per cent of those sixty and older.

Among all adults with high blood pressure, whether treated or not, 59.2 per cent knew they had it, 51.2 per cent were taking medication for it, and about 1 in 5, 20.7 per cent, had it below 130 over 80 — NCHS Data Brief No. 511, October 2024, published by the Centers for Disease Control and Prevention. A government data brief has no DOI, so the reference here is its series number.

Read those last three figures together, and note what they are a share of. Each one is a share of every adult with high blood pressure, not of the ones taking a tablet; the brief does not report control among treated adults on their own. What it does show is that about half of the people with high blood pressure are on a medicine for it and about one in five has a reading below 130 over 80. A reading that is still above target is not the exception. It is the ordinary case, which is why a second tablet is so common.

The second pill, and the ankles

When one medicine is not enough, one of the classes commonly added is the calcium channel blocker. It is widely used. It also has one side effect that is common enough to have its own meta-analysis, and it is the one in the photograph at the top of this page.

That analysis pooled 106 randomised trials and 99,469 participants. Swelling of the ankles and lower legs occurred in 10.7 per cent of people on a calcium channel blocker against 3.2 per cent on placebo or a comparison drug. It became commoner the longer the medicine was taken, reaching 24 per cent after six months. And it tracked the dose: 16.1 per cent on higher doses against 5.7 per cent on lower ones — Journal of Hypertension, 2011, DOI 10.1097/HJH.0b013e3283472643.

The mechanism in that paper matters for anybody about to cut their salt again. The swelling is thought to come from the small arteries of the leg widening while the veins do not, which raises the pressure inside the smallest vessels and pushes fluid out into the tissue. It is a plumbing effect of the medicine. It is not the body holding on to salt and water, which is why eating even less salt does little for it.

Three practical things follow, and all three belong in a conversation with the prescriber and nowhere else. The same analysis found the swelling much less common with some members of the class than with others. It found it was related to dose. And it found that more than one in twenty people stopped the medicine because of it over the long term, which is the outcome nobody wants. A tablet stopped without a word to the person who prescribed it is a far bigger problem than a tight pair of shoes. The swelling is a reason to ask. It is not a reason to stop.

When three medicines are not enough

There is a name for the far end of this road. The American Heart Association defines resistant hypertension as a reading above goal despite the concurrent use of 3 antihypertensive drug classes, taken at full or best-tolerated doses — Hypertension, 2018, DOI 10.1161/HYP.0000000000000084.

What is useful about that statement is the order in which it tells clinicians to look for the cause. Before anything else it asks for two checks: that the tablets are actually being taken, and that the high readings are not a white-coat effect, meaning a number that is high in the clinic and normal at home. Only then does it move on to diet, to other medicines that push the reading up, and to rarer causes.

The second of those checks is the one a household can do for itself, and it is cheap. A week of readings taken properly at the kitchen table is worth more to a prescriber than one reading taken in a waiting-room chair, and our page on reading a home blood pressure number goes through how to take one that can be trusted.

High in the morning, normal after lunch, high again at bedtime

The other thing people write down, and then stop trusting, is the log. Same arm, same chair, same cuff, and three different numbers in one day. The usual conclusion is that the machine is broken.

It almost never is. Blood pressure is not one number that the cuff either catches or misses. It moves through the day in everybody, and it moves again with the timing of a tablet, a coffee, a full bladder or a conversation held during the reading. A log that changes between breakfast and bedtime is a log that is working. What a prescriber wants from it is not one good reading or one bad one but the pattern across a week, taken the same way each time, and our page on reading a home blood pressure number sets out how to produce one that can be trusted.

The kitchen of 1958, and the question about plastic

One more subject turns up in the advertising around blood pressure, and readers ask about it often enough to deserve a straight answer. The argument runs that the American kitchen changed in the late 1950s, when cling film, foam cups and plastic tubs arrived, and that what came with them has something to do with arteries today.

Part of that is simply true. Kitchen plastics do shed, and how much has been measured: this site has gone through the published counts for cutting boards, containers reheated in a microwave and bottled water, with the limits of each study set out beside its numbers. What has not been shown is the last step. No trial has tested whether changing what a kitchen is made of lowers a blood pressure reading, and a page that says otherwise is ahead of the evidence. It is a reasonable thing to be curious about and a poor thing to be frightened by.

The supplement aisle

This is the point in the sequence where people start looking at capsules, and it is where the questions to this site come from. Somebody has cut the salt, takes two tablets, has ankles that swell, and sees an advertisement for a supplement that promises support for healthy blood pressure.

Some of the ingredients sold that way have been tested properly. Potassium and magnesium each lower the top number by two to four points in pooled placebo-controlled trials. Others have reviews that found little or nothing. And none of them has been shown to do what a prescribed medicine does, which is to reduce heart attacks and strokes over years.

Why this page carries an advertising link

Readers asked about one of those products by name often enough that saying nothing would have been its own kind of answer. So we did the work: opened the checkout, read the return policy line by line, went through the published reviews of the ingredients the category is built on, and wrote down what each one says, including the price and the parts of the checkout that cost money if clicked past.

That page is linked below. It carries an advertising link and so does this one, and if a reader follows it and buys something, this publication is paid a commission. Everything above was true before the link existed and stays true whether or not anybody follows it.

The questions people type

These are the questions that come up most on this subject, each with the short answer. The long ones are in the sections above.

• Why won’t blood pressure come down even after cutting salt? Because the average effect is about five points in people with high readings, and some people respond far less than that.
• Why won’t it come down after losing weight, or with exercise? Both lower it, by about one point per kilogram lost and three to eight points for regular endurance exercise. For a reading that started well above target, that can still leave a gap.
• How can someone eat salt every day and have normal blood pressure? Sensitivity to salt differs from person to person, and nothing a household can do will show which kind it is.
• Why are the readings never the same? Because blood pressure moves through the day in everybody. The pattern across a week is the information; a single reading is not.
• Can a scan show calcium in an artery wall? Yes. A CT calcium score counts it, and it is a separate thing from a cholesterol result. Whether to have one is a question for a clinician.

What we still don't know

Three questions, all open, and each one would change the practical answer if somebody settled it.

• Who responds to salt reduction and who does not. The average is five points in people with high readings; the spread around that average is wide, and there is no kitchen test for which side of it a person is on.
• Whether slowing the stiffening of large arteries, as opposed to lowering the reading, changes what happens to people over decades. The Framingham pattern describes the problem. It does not say what reverses it.
• Whether any supplement sold for blood pressure changes anything beyond the reading. The reviews measure a number over weeks or months, and stop there.

Swelling in one leg and not the other, swelling that comes with breathlessness or chest pain, or a home reading that is suddenly far higher than usual is a different subject from this page and belongs with a clinician the same day, not with a supplement.

General information about a common complaint and the published research on it. Not medical advice, and no substitute for asking somebody qualified who knows the rest of the picture. Nothing here is a reason to begin, stop or alter a prescription.