Here is my ‘bottom line’ on salt: The more you eat, the higher will be your urine calcium and – from what trial data I IMG_2303can find – the higher your risk of forming stones and losing bone mineral. The big articles, on salt and on idiopathic hypercalciuria have put into place the structure I need to make this simple generalization acceptable. That is part of the reason I wrote them – to be a structure I can rely on and work off of.

The large image is from a recent publication about the bone disease of idiopathic hypercalciuria. The full report is available and I recommend it

So here is a new kind of article, from me to you, a call to consider action on your own behalf.


Professor Loris Borghi and his colleagues, who performed the only prospective trial of reduced salt intake on actual stone formation wrote a fine review in 2014 that can be obtained in PDF form without charge. The review is a powerful argument for restraint, and well referenced.

Salt and Urine Calcium

The main point is that higher sodium intake raises urine calcium. The mechanisms are essentially that the way the nephrons get rid of salt force them to lose calcium more or less in due proportion.

male uca vs una Nephrol. Dial. TransplantHere is their graph of urine sodium – which faithfully mirrors sodium absorption from diet – on urine calcium in men. I have already introduced millimoles in thfemale Nephrol. Dial. Transplante sodium article so you can review the matter there. Women are the same. The graphs look different because the top of the vertical axis for men is higher than that for women, so the slope is higher for women. I would have prefered equivalent axes but take what I can get.

You already know about urine calcium and stone risk from the big hypercalciuria article: Above 200 mg/day is the beginning of increased risk. Take another look at these graphs. To get under that number you need to lower sodium intake below 100 mmol/day in men and about the same for women.

URINE CA VS NA WITH ALL DATA FROM ELAINE AND I COMBINEDI have made a graph just like theirs, from all of the data I could collect in which salt intake was deliberately varied in a trial setting and changes in urine calcium measured.

Here, prospective trials are circles, observational data are triangles. Blue points are from people with idiopathic hypercalciuria, red points are from normal people. This graph is perhaps more generalizable than those shown above because it comes from many separate trials from different laboratories.

It shows that IH involves an increased slope dependence of urine calcium on urine sodium, which latter reflects diet sodium intake. At urine sodium levels below 100 mEq/d the curves begin to overlap. The newest FDA guidelines for US people is an upper limit of 100 mEq/d and an optimal value of 64 mEq/day.

Curhan has already shown that in observational data more urine calcium means more risk of stones. All that is missing so far is to prove that less sodium leads to less stones, and that is what Borghi did some time ago.

Salt and Calcium Kidney Stones

Borghi’s trial was not strictly a lower salt trial because several changes were made at once, and it lacked some of the trappings of an ‘ideal’ trial in that no one can not know what they are eating – so double blinding about treatment is an absurd notion. But it was a trial of considerable importance and generated some insights that, I believe, are perhaps not fully appreciated.

What He Did and Why

My article on idiopathic hypercalciuria reviews the complex evolution of thought concerning pure ‘absorptive’ vs. ‘renal’ mechanisms and eventually comes to rest on a not unexpected consensus: High absorption and reduced renal calcium conservation both seem present. As a result people with IH will not in general – though there may be specific exceptions – tolerate low calcium diets without incurring risk of bone mineral loss. Until recently, in the absence of compelling evidence for a wide spread renal mechanism for hypercalciuria, low calcium diet was a common means for reducing calcium stone recurrence.

It is against that background that the Borghi trial needs to be analysed. The trial set out to compare ‘…the traditional low calcium diet with that of a diet containing a normal amount of calcium but reduced amounts of animal protein and salt’ in terms of new calcium stone formation in patients with IH – unfortunately only men. One stated reason for the trial was data from Curhan showing that high calcium diets seemed protective against stones – inverse statistical association between diet calcium intake and onset of stones (I have referenced his later paper (2004) which shows effects of age on the calcium – stone link; his earlier paper was in 1993). Another was the large body of data available even then linking high salt intake with high urine calcium excretion. A third was the established fact that low calcium diet can increase urine excretion of oxalate which would balance the fall in urine calcium so lower urine calcium would not effectively lower urine calcium oxalate supersaturation.

His diets were of Italian cuisine and therefore not necessarily transferable to the US. But the key compositional differences are certainly simple enough: High calcium low sodium reduced protein – Calcium 30 mmol (1,200 mg), sodium 50 mmol (1150 mg sodium), protein 93 gm. Low calcium diet – Calcium 10 mmol (400 mg), avoidance of highest oxalate foods (walnuts, spinach, rhubarb, parsley and chocolate). Both diets included advice to maintain good hydration.

His male patients all had IH using the research criterion of 300 mg of urine calcium daily and at least two documented kidney stone passage events in the absence of any named systemic cause of stones or hypercalciuria. From 354 candidate cases, they excluded 234 and were left with 120, 60 in each of the two diet groups into which patients were enrolled using a random allocation system. Both groups were eating a lot of salt before the trial – above 200 mmol (4600 mg) daily.

What He Found and Why

borghi trial famous figure


This famous graph simply shows the life tables of the two groups. After about 2 years, the normal calcium, low sodium reduced protein diet patients formed many less stones than did those eating the ‘traditional low calcium diet’.

The numbers in the table below the graph show how many people were ‘at risk’ over time – made it that far into the trial and did not drop out. They are comparable, and adequate for the one big conclusion: Low sodium high calcium reduced protein diets work better than low calcium diets and given what we know know about IH are a lot safer for bones.


But why did the numbers of stones fall.

Supersaturation. Everything on this site should point to a greater fall in supersaturation as the mechanism, so let’s look there. Calcium oxalate supersaturations (rounded without variance estimates) were 10, 8.1, 7.3, 6.8, 6.7, 4.8, and 4.7 for baseline, 1 week, and years 1-5 inclusive in the low calcium diet group. Corresponding values for the reduced sodium, protein and high calcium diet were 9.6, 4, 5.1, 4.7, 4.5, 3.7, and 3.5. Bolded means lower than low calcium diet by a variety of comparisons. One would have liked the calcium phosphate supersaturations given that they may be even more important than those for calcium oxalate, but we do not have them.

Urine Calcium. The low calcium diet and the high calcium reduced sodium and protein diet gave about the same levels of urine calcium, about 6 – 7 mmol (240 – 280 mg) daily with no differences between the groups.

Urine Oxalate. Because high calcium intakes reduce urine oxalate, and low calcium intakes raise it, the low calcium diet group consistantly excreted more oxalate: baseline, 1 week, and years 1-5 inclusive: 357, 422, 422, 411, 422, 433, 411 mmol/day vs. 411, 322, 344, 322, 333, 333, 333 mmol for the lower sodium high calcium reduced protein group – bolded means significantly lower. It is this, the lower oxalate excretion, that lowered the calcium oxalate supersaturation; urine volumes were the same.

Urine sodium. How well did people follow the diet? As expected, the low calcium diet group continued to lose above 200 mmol/day of sodium. The low sodium group with a goal of 50 mmol/day actually excreted between 110 and 130 mmol, except for week one which was indeed 51 mmol/day. So the outcome was double the goal. The urine calcium was, however, no higher despite the high calcium intake, meaning that the low sodium overcame the effects of the higher intake.

What I say

The trick is this: If sodium and protein are controlled one can eat a lot of calcium – 1200 mg – and yet keep urine calcium oxalate supersaturation reasonably low because the high calcium intake will lower urine oxalate and the low sodium will lower urine calcium.

The protein effect is not simple. Protein imposes an acid load and was thought to raise urine calcium because of the acid loading which would mobilize bone mineral. But more modern studies have shown this is not the case. Protein increases calcium absorption and urine calcium excretion even when enough alkali is given to offset the acid production from protein. A decent literature indicates that protein probably promotes bone mineral retention, not the opposite.

I have deliberately not referenced this last paragraph as I mean to deal with the protein effect in another article and do not want to bog us down.

Let us say that 93 grams of protein would be about 1 – 1.2 gm/kilogram body weight, not really very low. They used more from milk, breads and pastas than US people usually do, but the Italian cuisine favors that approach.

Perhaps for the moment I can say – though I will be criticized – that if we achieve reduced salt intakes of about 100 mmol (2300 mg sodium) daily and keep protein intake around 0.9 – 1.1 g/kg (all routine 24 hour kidney stone urine tests report a protein catabolic rate (PCR) value which gives an estimate of protein intake in gm/kg) we can eat 1200 mg of calcium in the diet and achieve protection against new calcium stones if we are men with IH. A women study is important here, and one in the US.


I already reviewed the bone story in kidney stone formers. I know my article is long and difficult but it is accurate and brings together material from other sources that are also difficult.

Bone disease follows stone disease like a shadow. It is hidden among our patients until the first fractures. Often we can be so focused on the stones we forget about the silent but perhaps even more serious problems that are developing in bone and will become obvious enough years later.

We not only do not have bone oriented trials for stone formers, we do not even have a deep enough knowledge of the mechanisms producing bone disease, which is probably in large part from idiopathic hypercalciuria and from intentional or unintentional reduction of diet calcium intake combined with high sodium diets.

But at least one really wonderful bone trial points to low diet sodium and high diet calcium as a promising treatment to maintain bone mineral balance in IH and perhaps prevent fractures.

Bone Balances

There is every reason why reduced salt intake coupled with high calcium intake should benefit bone mineral balance. Trials are few indeed. This one is so elegant and convincing I have selected it.

Birgit Teucher pointed out in 2003 that salt loading should impair bone mineral balance, and that direct proof of such a link was wanting in humans. Her review is worth a careful read, as to me it seems a preliminary to her massive balance study in women testing the effects of high and low calcium and salt intakes. If her review is adequate, as I believe, her balance study seems to stand alone up to its publication, and to the present time – as best I can tell.

In my last paragraph and the following analysis of the main research I refer to Teucher as if she did the work alone. I mean no disrespect to her many coauthors but detest the passive voice and the tedious refrain of ‘and colleagues’.

Her own study is a crossover randomized trial of low and high sodium (70 and 191 mmol) and low and high calcium (518 and 1284 mg/day) calcium diets in a square design: low calcium and sodium, high calcium and sodium, low calcium high sodium and high calcium low sodium, each consumed for 40 days with at least 4 weeks washout in between. Sixteen postmenopausal volunteers entered the study and 11 finished all four dietary interventions.

The way she did the diets is of interest, to me at least. She constructed a ‘basal’ low sodium low calcium diet to which she added amounts of sodium in meal preparation and as supplements, and calcium exclusively as calcium carbonate. The supplements were taken with the meals, so far as I can tell.

She measured calcium balance two ways. Urine and fecal calcium were measured directly on days 30-40 of each meal cycle, and could return a classical estimate of overall systemic calcium balance. Her key approach went further. She used a complex calculation made from two stable calcium isotopes, one (calcium 42) orally the other (calcium 43) intravenously. This latter double isotope technique permits an estimate of net calcium absorption and bone mineral balance.

PQ BAR CHART OF ABSORPTION ENDOFECAL URINE AND NET BONE BALANCE BY HI LO CA AND NAShe published her results in a group of tables, but I have graphed the main findings for clarity here. My source is her Table 2.

The vertical axis shows calcium in mg/day. The horizontal axis shows the four diet periods.

The crosshatched blue bars show calcium absorbed from diet. It is lower, of course, with the low calcium diet.

The next bar to the right is secretion of calcium from blood back into the intestinal lumen by pancreas, liver, and other transporters; it is a net loss of absorbed calcium from the body and can be determined only using the isotope studies.

The third bar, counting from the left, is simple urine calcium excretion, which is loss from the body.

The fourth bar is net bone mineral balance.

Positive balance occurred only with the high calcium low sodium diet and resulted from a fall in endogenous fecal (endofec) calcium secretion and urine calcium combined, even though in her statistical analysis only the urine loss difference was significant between high and low sodium periods on the high calcium diet.

Implications for Present Stone Treatments

Does this mean that stone formers with IH who restrict their daily sodium intakes below 100 mmol (2300 mg) and add calcium supplements to their diets to achieve 1,200 mg/day will avoid bone disease?

No, it does not.

We have no comparable study on female hypercalciuric stone formers and no study of men. Notice how low were the urine calcium levels in her women – 112 and 153 on the low and high sodium, high calcium diets. Perhaps hypercalciuric stone formers would benefit greatly, perhaps not. Perhaps men, hypercalciuric or not, would behave differently. Likewise for premenopausal women with IH vs. those who are postmenopausal.

But – that word again! – reduced sodium high calcium diet seems to protect at least Italian men with IH from more calcium stones. It is a good diet alternative. And, it just so happens it is more or less the same diet that Teucher used for her women.

It is also in line with modern general health recommendations. I have already referenced the CDC and American Heart recommendations for reduced diet sodiumThe diet calcium recommendation of 1,200 mg is well known

We need more trials. But while those of us on the sidelines wait, I think low sodium high calcium diets are right for IH stone formers, men and women, and will use them, as I do use them, with enthusiasm and persistence.

As a practical point, Teicher taught us that supplements can be taken with meals and the calcium will go into bone. So perhaps it is not so critical that all the calcium comes from food alone.

Implications for Those Who Fund and Perform Stone Trials

What We Need

We need more trials like her trial. How about IH men and women?

We need more trials like his trial. How about women, how about a US cuisine?

Teucher used supplements with meals in place of food calcium and got good bone results. What happens with IH stone formers? Is this not  grounds for a good trial?

It is diet sodium, calcium, and protein that we need to sort out, following the wonderful work by Borghi and Teicher.

What We Do Not Need

Why nibble at straw when surrounded by sweet delights?

Need we prove again that thiazide or potassium citrate work? I say no.

Do we need to test again if high fluids are a good thing for stone prevention? I say perhaps, but with minimal enthusiasm.

Borghi did a fine trial of fluids years ago. It was among the first items I put up on this site. In a randomized but of course not double blind trial, 12 of 99 first time calcium oxalate stone formers with a daily urine volume averaging 2.6 liters formed a second stone over 5 years vs. 27 of 100 well matched patients whose urine volume averaged about 1.1 liters. If we were to do this again, how likely is it the results would be different?

What Might Hold Us Back from Really Worthwhile Trials?

Why not tackle the the less well studied issue of diet sodium, calcium, and protein in US men and women? It bears not only on stones but on bone disease, and is alternative and in addition to fluids which, quite frankly, every reasonable physician freely recommends because without cost and without risk.

I have thoughts about why diet sodium and protein, and effects on bone, have not been studied.

What Borghi and Teucher did is hard to do.

Her study was heroic.

His was long, arduous, and risky – who knew what might be the final answer.

We cannot redo his study as he did it. No one would use low calcium diet as a contrast anymore, so the contrast would have to be an active drug like thiazide or potassium citrate.

Both studies were done outside the US; our cuisine is distinctive and perhaps US citizens might not adhere to diets as well as Italians.

Is it not time to climb higher, breathe a thinner if more invigorating air?


Scroll to Top