MeUp to this point we have considered only increase of urine volume as a means of stone prevention. The effect of increased urine volume is to reduce urine supersaturation with respect to stone forming salts and therefore reduce the risk of crystal formation which is the basis for kidney stones.



Supersaturation with respect to the calcium stones depends upon urine concentrations of calcium, oxalate, phosphate, and citrate, and, in the case of calcium phosphate stones, or uric acid stones, urine pH. Giving citrate salts can reduce urine calcium excretion and increase urine citrate. Urine citrate binds urine calcium in a soluble citrate complex, which reduced calcium salt supersaturations. Citrate inhibits crystal formation, growth and aggregation. The alkaline citrate salts can raise urine pH.

relative risk vs urine citrate from Curhan control file in stone bookEpidemiology

In a prospective study of two nurse (red) and one male physician cohort (blue) Curhan found that relative risk of kidney stone onset (vertical axis) rose as urine citrate excretion (shown in hexiles along the horizontal axis) fell. Below 400 mg/day of urine citrate risk was – compared to above 800 mg/day) increased by nearly 2 fold. Mean relative risk is at the ends of the shaded bars. The upper 95% of risk is at the tops of the filled bars. Even though the average risk (end of crosshatched bars) remained below 1.



Although I had quibbles with some of the comments it included, I believe the recent American College of Physicians (ACP) review of kidney stone prevention trials was done properly, and therefore have selected for review here those they felt were technically adequate.

Below is a detailed presentation of the five studies. Here is a link to my spreadsheet with all of the numbers. It also contains my references for thiazide treatment.

Ettinger et al (J Urol 158:2069-2073, 1997).

Sixty four patients with at least 2 stones in the past 5 years and at least 1 within the past year before the trial were given placebo (33 cases) or potassium magnesium citrate (63 mEq citrate, 42 mEq as potassium and 21 mEq as the magnesium salt in combination pills) – 31 cases. Each pill contained 21 mEq of citrate; 2 pills were taken 3 times a day. The trial was designed to last for 3 years. There were 5 and 9 women in the placebo and treatment arms. Urine citrate excretions were not different before treatment (549 and 587 mg/day, respectively, nor were urine volume, pH, calcium, oxalate, or any other stone forming risk. After a one month grace period in which new stones were not counted, any passage or radiographic appearance of new stones, or growth of previous stones was considered a treatment failure. During the trial, 15 subjects left the treatment arm, 8 the placebo arm.

New stones or growth of old stones occurred in 63.6% (16 cases) of the 25 placebo cases who finished the trial and in 12.9% (2 cases) of the 16 treated cases who finished the trial. If the 6 subjects who left the treatment arm because of drug adverse effects are added in as treatment failures the drug effect remained significant (8 of 22 or 33%).

Of note, this particular formulation is not available in the US. A version of the supplement is available OTC but the dose per pill is so low that it is impractical for anyone to use it. So the trial is part of a proof of principle, but not actually applicable to clinical practice in this country.

Lojanapiwat et al (International Braz J Urol 37:611-616, 2011)

Unlike the Ettinger study, which concerned spontaneous stone formation, this study concerned new stones or growth of residual fragments after shock wave lithotripsy (SWL) or percutaneous nephrolithotomy (PERC). Their subjects were 80 initial patients, all 8 weeks after either procedure, and either stone free or having no residual stone fragments >4mm diameter (Numbers are in the Table). Hypocitraturia (<325 mg/day) was present in 20/39 who received citrate and 15/37 who did not.

  SWL PERC  Total
STONE FREE 24(8)  15(5)  39(13)
RESIDUAL STONES 26(17)  11(9) 37(26)
Total 50(25)  26(14)  76(39)

They were randomized into 39 treated and 37 placebo treated groups and followed for one year which 76 of the original 80 completed. Numbers receiving citrate in each group are in parentheses. Sodium potassium citrate was given as 81 mEq/day in 3 divided doses).

Of the 13 cases who were stone free and received citrate, 12 remained so vs. 15 of the 26 given placebo. Of the 26 who had retained fragments and were given citrate, 8 were stone free vs. 1 of the 11 controls and 16 others given citrate showed no change (13) or reduction in size (3) vs. 2, no change and 2 decreased size among the 11 placebo. These differences were judged significant at the p<0.05 level by the authors.

Soygur et al (J Endourology 16:149, 2002)

This trial considered 90 patients after SWL for lower pole stones who had residual stones <5 mm or were stone free. They were randomly assigned to potassium citrate (50 mEq/day in 3 divided doses) or placebo (Table). The trial lasted one year. The end

  Citrate  Placebo  Total
STONE FREE 28(0)  28(8)  56(8)
RESIDUAL STONES 18(0)  16(6) 34(6)
Total  46(0)   44(14)   90(14)

points were stone free or not and residual stone size increased or not.

New stones occurred (parentheses) in none of the citrate treated stone free patients and in 8 of the placebo treated patients. Among the residual stone group, the fragments disappeared in 8 treated cases and failed to grow or shrank in the others vs. growth or new stones in 6/16 placebo cases. The differences in growth or new appearance were all significant.

Of course, both of these post treatment trials are subject to the biases of a radiography study, but observers appeared to have been suitably blinded to the patient groups.

Hofbauer et al (British J Urol 73:362-365, 1994)

In this trial, an equimolal sodium / potassium citrate was given in doses that maintained urine pH in the range of 7 to 7.2 vs. placebo. Therefore, although patients were allocated randomly to active treatment or placebo, the trial could not be blinded. By the three year endpoint, 22/25 placebo and 16/25 active drug subjects remained. New stones occurred in 16/22 placebo and 10/16 active drug subjects. This difference was not significant. This study is the only one with a negative outcome. It is also the only study that was not double blinded.

Barcello et al (J Urol 150:1761, 1993)

Stone formers with urine citrate excretion rates below 643 mg/day (3.4 mmol/day) were allocated to potassium citrate 60 mEq/day in 3 divided doses. Their mean urine citrate excretion was 359 mg/day. At the end of three years of followup, 20/28 placebo treated and 18/27 citrate treated subjects remained. New stones occurred in 14/20 placebo and 5/18 treated cases, a significant departure from chance.


TREATED  20  115 135 
NOT TREATED 77 71  148
TOTAL  97  186 283 

Despite the variability of design, one can, with nerve, simply ask about the beneficial effects of citrate salts across all the trials. In all five trials 283 people completed the desired treatment period. Of these, 97/283 (34%) formed new stones or, in the case of the post procedure trials showed growth of retained fragments. Among all patients who were given citrate salts, 20/135 (14.8%) formed new stones or showed growth of retained fragments vs. 77/148 (52%) of those given placebo.

I have not added back the 6 cases from the Ettinger trial who left because of drug side effects.

From this we can reconstruct a sense of the value of the treatment as applied to the mixed practice of post surgical management and overall medical prevention.

Let us assume these numbers will hold for the future.

For every 1000 cases like the ones in the trials, 520 untreated cases will form new stones or show stone growth after a procedure vs. 148 cases/1000 cases with citrate, a savings of 372/1000 cases overall.

I realize I am not calculating in the most satisfactory manner as a statistician, but I rather like the coarse grained, even vulgar nature of my count me up.


A Personal View

The trial community exhibits the kind of methodological fussiness one expects and applauds in any scientific situation. Among their ilk the citrate effect is viewed as modest at best, the evidence, by their likes, fair.

I am sure they are right according to the mores and social instincts of this discipline, but I do not come from nor inhabit that discipline, and therefore have an altogether different way of counting – for that is all one does after the impatient and often indifferent subjects have played out their roles in the work.

How likely is it, I ask myself, that citrate salts do not prevent new stones or fragment growth?

Not at all likely.

Why assume anything but that blinding was performed when specified, that radiograph readers were competent and blinded to the groups patients were in, that stone events were counted fairly and compared to radiographs to estimate new stones? If we make these assumption of honesty and skill, the marked downward skew from alkali is just too large to be by chance.

My bet will be on the drug, and if I bet that way, I will always win.

Do We Need More Trials For Calcium Stone Formers?

For me, no. It would seem a waste of money.

Some trials treated patients with reduced urine citrate, others did not. Some trials looked at new stones over 3 years, others at residual fragment growth one year after urological procedures. Will another 50 or even 100 cases be likely to change the outcomes? If so, in what way, and why?

It is true that one trial showed no effect and that trial was not blinded. It is actually a drag on the results as I did not remove it.

We Do Need a Trial of Citrate for Calcium Phosphate Stone Formers?

I do not know how often this must be said. Calcium phosphate stone formers must lurk in each of the trials I have reviewed, but I do not know their outcomes. One trial insisted stones be at least >50% calcium oxalate. That means perhaps a few had considerable phosphate is stones.

Calcium phosphate crystal formation is sensitive to urine pH whereas calcium oxalate stone formation will not be. The reason is that calcium phosphate supersaturation requires divalent phosphate be present, and the pKa for the second proton is about 6.8. Citrate salts can raise urine pH, so they can raise supersaturation with respect to calcium phosphate salts. On the other hand, citrate is an inhibitor of crystallization both because it is calcium binding and because it directly affects calcium crystal growth.


The very same ACP report from which I derived the studies shown here presented an annoying set of comments that infers we might as well just give a drug like potassium citrate without knowing stone composition, or doing serum or urine testing that concerns stone pathogenesis.

For this reason, I offer some remarks on that subject. This is in the special context of citrate treatment. I have made more general remarks of a negative sort about the APC comments.

Does Stone Analysis Matter?

How can it not? I have already mentioned the problem of phosphate stones. Do we not have to exclude struvite is stones? The odd patient with cystinuria who has slipped by? Drug stones? Conversion from calcium oxalate to calcium phosphate stones?

Do Serum and Urine Testing Matter?

How can they not?


Do we want to give potassium loads to people with reduced renal function?

Having prescribed potassium, do we not want to monitor for serious increase in serum potassium; some patients are older, some diabetic, some take ACE or ARB medications, some age or change drugs over the years we treat them.

Do we not want to diagnose primary hyperparathyroidism? You cannot without serum testing and 24 hour urine testing to be sure calcium excretion is not low.


If we do not obtain and measure 24 hour urine samples, how can we know anything? Some patients may have very high urine citrate levels. Some may have very high urine pH values.

Here and there urine oxalate is very high, from primary hyperoxaluria, or occult malabsorption syndromes, or very odd food habits.

People change their habits and develop diseases.

Moreover, people do not always take their citrate. Fall in urine ammonia in relation to urine sulfate, and rise in urine potassium assure one they are taking the drug.


Do We Need a Trial for Uric Acid Stone Formers?


No one really questions that alkali salts will raise urine pH, nor that raising urine pH will reduce uric acid supersaturation and prevent stones. It is common practice. I doubt anyone will pay for or perform an RCT to test this question.

That they will not is very important, because it raises an unexpected question.


We Know the Chemistry

Uric acid is a large flat mainly hydrophobic molecule with most of its charge on a single proton receptor site. The protonated from has a very low solubility in urine of around 90 mg/liter whereas 24 hour urine uric acid excretion ranges from 400 to over 1000 mg daily depending upon diet purine loads. The pKa of the proton receptor site is about 5.3 in urine. 

Given these facts we can calculate uric acid supersaturation from the urine concentration of total uric acid and the pH, along with minor adjustments for the effects of ionic strength on the pKa. High supersaturation will lead to a snowstorm of uric acid crystals. Raising urine pH to above 6 will generally reduce supersaturation below 1 and end uric acid stone formation.

Everyone Knows Alkali Work

There is a lot of uric acid excreted every day, so uric acid stones can grow rapidly. Uric acid gravel has an orange red color and is often seen. When alkali are given, the gravel goes away only to come back if patients miss doses. The absence of new stones is obvious.

No One Treats Without Stone Analyses

Who can be sure of stone composition without stone analysis? Even during treatment of someone who has produced uric acid stones, calcium oxalate or calcium phosphate stones may begin. So people know the stone type, and proceed by custom.

No One Treats Without Testing Serum and Urine

Uric acid stones are common in diabetics and people with reduced renal function; potassium loads are potentially dangerous. Perhaps this is more obvious among uric acid stone formers than calcium stone formers, although given wide spread use of ACE and ARB drugs and NSAIDS, potential risk is everywhere.

The amounts of alkali needed can be variable, and the only reliable way to ascertain is 24 hour urine testing. Likewise for compliance.

Therefore routine practice monitors before and during potassium citrate treatment of uric acid stones.


IN this situation, no one has and probably no one will propose a trial of alkali for uric acid stones. But, there is an almost exact parallel situation for calcium phosphate stones, yet such certainty as pertains to uric acid stones certainly does not exist.


Do We Need a Trial for CaP Stone Formers?



We Know the Chemistry

Calcium cannot combine with mono-valent phosphate but only with the divalent form. The pKa for dissociation of the second proton of phosphoric acid is about 6.8 in urine, although the precise value varies with ionic strength. Given the molarities of total phosphate, calcium, citrate – which binds calcium – and other ligands that have modest effects, the supersaturation of brushite – the usual initial urine CaP phase – can be calculated as well as we can calculate the supersaturation for uric acid.

Like uric acid, phosphate and calcium are abundant in urine, so the amount of crystal that can be produced in a day is similar to that of uric acid. Therefore stones can, and do, form rapidly and become large.

As in the case of uric acid, high urine CaP supersaturation can produce snows storms of crystallization; though certainly not common, patients can recognize this as white urine.

On physical chemical grounds, to lower CaP supersaturation below one and keep it there is to prevent CaP stones as surely as one prevents uric acid stones by raising urine pH and lowering supersaturation below one. Why, then, is not this treatment as self evident as alkali for uric acid stones?

Everyone Does Not ‘Know’ Treatment Works

We have no drug corresponding to alkali.

We can raise urine pH safely but cannot lower it.

Acid loads raise urine calcium losses and can be detrimental to bone mineral balance. Higher protein intake is a possible way to lower pH, but not all kidneys respond to acid with a prompt fall in pH. In some cases urine ammonium ion excretion will rise. In others, acid retention may occur. Urine calcium will tend to rise.

So treatment is not as transparent as for uric acid.

But Treatment Must Work Exactly the Same Way

We can lower CaP below 1 with fluids and measures – reduced diet sodium and thiazide – that reduce urine calcium, and we can monitor supersaturation as we monitor urine pH and uric acid supersaturation.

Furthermore, patients can tell if white urine has ceased.

Moreover, because stones are often actively forming, effective treatment is reasonably obvious.

However, these measures may be difficult to achieve. Thiazide is not always tolerated, reduced salt diet not always maintained.

Citrate is a powerful inhibitor of crystals, and it would be good to know if it were beneficial for the CaP stone former.





  1. Rasmus

    Dear Dr. Coe, many thanks for a wealth of information. Potassium citrate is mentioned frequently, would magnesium citrate also suffice?

  2. Lisa M Viviano

    Dr. Coe,

    Thank you for your continuing efforts to help with identifying the causes of renal calculi and minimize kidney stone formation. I began following your recommendations last summer and finally saw a reduction in my stones! Now? I am struggling with a compromised medication that caused my stones to return. Long story I won’t bore you with, but, I am at least sure the regiment you recommend is effective. It is difficult for me as I had been on a diabetic diet and so many of the foods I had learned to love are high in oxalate. I’m adjusting my diet, again, as the stones are a much greater risk to my health than a slightly elevated A1C. I will see my nephrologist soon and ask about a Rx for Urocit K. I’m hoping that will further reduce my stones.

  3. Jim Russell

    Many thanks for reassuring me that the increased water/increased calcium/low oxalate regimen I’m on makes sense. It means a lot to know that it is likely doing some good since I’ve had to significantly change drinking and eating habits Thanks again for generously sharing your knowledge and time.

  4. Luke

    Hello again.

    Since beginning of Septeber, I am measuring my urine pH each time. What makes me curious is the fact that in the morning pH is very low – and despite citrate load (10 mEq + 0.5 dm3 of water and 6 am and 8 am ) it stays below 6.0. After first meal (~12am) it rises to ~6.5. Previously, when I was eating a small breakfast (+ 10 mEq Kcit) at 6am, it was always about 6.5. However, if I don’t eat much protein meals, then pH falls @ 5-6pm (another doses of Kcit @ 2,4 and 6pm) to 5.8-6.0…
    If I eat a high protein meal ~5 pm , then pH skyrockets to 7.2-7.4 about 2-3 hours after (even if I stop citrate) and stays there for a few hours , to my last visit in WC before sleep. If I eat heavy meal before sleep (bad idea, I know) – then even in the morning pH is above 7…

    So, for now, since I cant consult my doctor…I am trying to go for low amount of food distributed through all day and that works… but I never suspected our meals influence urinary pH so heavily.


    • Fredric Coe, MD

      Hi Luke, you are a good scientist. Yes, meals raise urine pH – well known, and you have accurately observed it. For uric acid stones I would not be too fussy, it is the 24 hour urine pH average that counts, so just use enough potassium citrate or other alkali sources to get it above 6. Regards, Fred Coe

  5. Jim Russell

    Dear Dr. Coe – How much of your advice applies to bladder calcium oxalate stone formers? I have had three procedures in the last four years to remove them. I have BPH but very low residual after voiding. This had led my current urologist to think the cause is metabolic rather than BPH-caused obstruction. I’ve had a 24 hour test which showed 1.8 liters and high oxalate. Treatment now consists of increasing my water intake to eight 8 oz. glasses of water daily and eliminating high oxalate foods. Does this seem reasonable to you? Many thanks in advance for your advice and work.

    • Fredric Coe, MD

      Hi Jim, Your doctor may be right. The way to proceed is not to add fluids and fool around with oxalate. Get things evaluated the right way and treat what is abnormal. Try this plan and see if it works for you. Regards, Fred Coe

      • Jim Russell

        “The way to proceed is not to add fluids and fool around with oxalate.” Is this correct? Did you mean now instead of not?

        • Fredric Coe, MD

          Hi Jim, Thanks. I cannot find this phrase in the citrate article. Can you point to it? I read the article through 3 times and seem to have missed it. Regards, Fred

          • Jim Russell

            It’s your second sentence in your initial response to my question.

            • Fredric Coe, MD

              Hi Jim, Yes I meant ‘not’ because of what follows – you need to get a complete evaluation to determine what to do and not try things that have been proposed for stones. Empirical treatment without chemistries is not efficient and will not in the long run work well. I pointed you to a good article with a decent sequence to follow. Regards, Fred Coe

              • Jim Russell

                The treatment plan to increase fluid and decrease oxalate was based on a 24 hour collection and analysis with the following results: volume 1.8; SS CaOx 6.19; calcium 111; oxalate 49; citrate 714; SS CaP .43; ph 5.809; SS Uric Acid .92; and Uric Acid (g/day) .578. The volume, SS CaOx and oxalate measurements stood out. Is there another test I should be taking? My urologist’s plan is to have an ultrasound test in March to see if bladder stones have formed. I greatly appreciate your willingness to respond to questions from the great expanse of the internet. Jim

              • Fredric Coe, MD

                Hi Jim, A reasonable plan. Be sure you are getting the full 1,000 to 1,200 mg of diet calcium; taken with main meals high calcium foods block oxalate absorption and make the diet issues a lot easier. The calcium also protects your bones.

              • Charlene

                In 2013 I was diagnosed with a renal oncocytoma and 1.7 centimeter stone. A blockage, a stent and a lithotripsy followed. After a 2nd lithotripsy, another blockage and then a 3rd lithotripsy, I was found to have just a few very small stones. I followed a low oxalate diet for nearly a year and asked my urologist about Theralith. I passed the fragments and have been stone free since March of 2014. I also have gout, but haven’t had a flare till this summer. My urologist told me to add foods back into my diet and drink at
                least a gallon of fluid daily and make part of that lemonade. Just wanted your opinion. Thanks for your time!

              • Fredric Coe, MD

                Hi CHarlene, Thanks for sharing your experience. Theralith is a low dose formulation of magnesium potassium citrate. This agent was tried on one trial but with higher amounts than I suspect you used with this product. Of importance, you did not say what your stones were made of. Given gout, perhaps they are uric acid. Regards, Fred Coe

Leave a Reply