Thiazide Type Diuretics Reduce Stone Formation
The common thiazide type drugs in use today are hydrochlorothiazide, chlorthalidone, and indapamide. All three have been used in stone prevention trials and shown to have beneficial effects. A nice recent review is also the source for the drug structures pictured above.
I have obtained and studied nine trials. In all nine trials, there was a comparison – untreated – group. This spreadsheet contains links to the trial documents, but you will find it not easy to obtain the original publications unless you have access to a university library system. For this reason I have copied out the key data. Briefly, there were 330 controls, of which 149 relapsed (45%), and 314 treated subjects of whom 72 relapsed (23%), a savings of about half (23%/45%). That is the bottom line for this class of drugs.
Here is a picture of the 9 studies. For each study the left panel has a bar whose height is how many control – red, and treated – blue subjects it had. The right panel shows how many of the red and blue people made new stones.
It is obvious that there were no differences in new stones between control and treated people (right panel) in studies 1 and 6. These were both brief (see the spreadsheet). In the others, the treated – blue bars – were lower than the red controls. Trial 9 had no relapses in the treated group. (Trial 5 has two drug doses and one control group so the height of the blue bar for 5 on the left panel is the average of the number of subjects in the two dose groups).
By now many of you will want some statistics. The spreadsheet depicts 9 studies each with numbers of treated and untreated people who did and did not relapse. Using ANOVA in which the numbers who relapsed is dependent, treatment or not is a categorical variable, and numbers of total subjects is a covariate, the mean number of relapse subjects adjusted for number of subjects was 16.1 in the control group and 8.3 in the treated group (p=0.008).
A simple X2 test using the four numbers from the bottom line of the spreadsheet: Treated, 72 relapse 242 no relapse; untreated 149 relapse, 181 no relapse gives a X2 of 35.2, p<0.001 (Yates’ correction is virtually identical for those who care).
I have deliberately included all studies that had a control and treatment group without fussing about the ‘quality’ of the trials. Several expert groups have reviewed these trials and a few more I omitted. Fink and colleagues in 2012 reviewed all available stone treatment trials. The same group did a similar review for the American College of Physicians which I did not think much of because of its codicils that seemed inclined to mediocrity – if not folly in practice. Escribano and colleagues authored an outstanding Cochrane review focused on thiazide.
The experts properly quibble about the intricacies of trial conduct and call for more trials. Personally I think more thiazide trials are of marginal importance and little interest. Mine is the untutored opinion of a bystander who does not perform trials as a profession. But I can count, and no one can tell me that the main fact – the pills reduce stone recurrence – will be overturned by trials of increasing ‘quality’. That is very unlikely. So, why do them?
Thiazide is Not Used Alone
Because they can reduce formation of new calcium stones, thiazide type diuretics are co-equal with potassium citrate as a medication physicians can use for stone prevention. Thiazide works differently from potassium citrate, so the two drugs can be used together with additive effects.
But more to the point, diet and lifestyle changes are crucial for stone prevention and need always put in place before writing a prescription.
The goal is to lower supersaturation. Fluids will do this. So will reduced diet oxalate, or raising urine citrate and, in uric acid stone disease, urine pH. Reduced sodium diet will lower urine calcium, and that will lower calcium salt supersaturations. So fluids and diet come first. Aspects of living that promote dehydration are serious obstacles to prevention, and to learn what they are in a given person and some to ways around them is an art of considerable passion and value.
Trials are all about the effect of a pill, and that is a virtue. But no doctor who practices stone prevention with a mere pill can achieves the joyous satisfaction of an accomplished clinician. So all I want or need from trials is to know the pill can reduce stones – presumably by reducing urine calcium, and thereby feel justified to add it into my instruments of practice, which I had done decades ago.
Why Use Diuretics?
It is Not to Raise Urine Volume
Because of the name one might think the drugs work by increasing urine volume, and therefore prevent stones the way water or other fluids might. That is not the case. Diuretics raise urine volume only transiently. Urine volume is set by how much extra water is available for urine loss. After a short while on any diuretic, the average day’s urine volume will be no different taking a thiazide than it would be not taking it.
They Lower Urine Calcium and Urine pH
How they do this is a matter for some detailed commentary, but let us begin by saying they do indeed. When physiologists comment on how this occurs the correct answer is that they lower the urine calcium by stimulating the kidney tubule cells to reclaim back into the blood a higher than usual fraction of the calcium that the kidneys filter out of the blood in the usual course of their functioning. How they lower urine pH is not well known.
But higher calcium reabsorption, though true, could not lower urine calcium for very long any more than reducing the radius of a bathtub drain can lower the flow of water out of the tub whose faucets are open. If you did that, narrow the drain of such a tub, the water would rise until the greater weight of the water column raised outflow back to match the inflow from the faucets or the tub simply overflowed onto the bathroom floor. The amount of calcium in the urine every day is exactly the amount entering the blood from diet and bone. So thiazides have to somehow alter that inflow – either reduce diet calcium absorption or promote calcium entry into bone. It is mainly the latter they do, so they are not without some potential to maintain bone health.
They Must Lower Calcium Salt Supersaturations on Average
This is an old theme on the site. Reduction of supersaturation will lower formation of stone crystals and eventually stones. Given a random variability of urine volumes, and of other key factors that control supersaturations for calcium oxalate and calcium phosphate – eg. excretions of oxalate, phosphate, citrate, and urine pH, a fall in urine calcium from a drug will inevitably lower supersaturations on average unless that drug systematically raises urine oxalate or phosphate excretions, lowers citrate excretion, or alters pH upwards – calcium phosphate. In fact, thiazides can lower urine citrate, probably in part because they deplete body potassium stores and lower the pH inside kidney cells, but that can be corrected by potassium repletion. Lower pH from thiazide will specifically reduce calcium phosphate supersaturation.
How do Thiazides Affect the Kidneys?
Calcium and Sodium
With a brief blush and downward gaze, I choose our own publication as perhaps not unreasonable as a source. Partly I am guilty of favoritism, partly we are the only group to have studied individual stone formers before and during thiazide treatment in a clinical research setting on a constant diet. We did do this, and the results, if perhaps not exactly arising from a multitude, are secure in their precision and in their depth of insights.
Four men with calcium stones and idiopathic hypercalciuria were studied in our CRC before (gray symbols) and after 6 months (black symbols) of chlorthalidone, 25 mg daily used in treatment for stone prevention.
We have spoken of filtration and reabsorption elsewhere on this site. The drug had no effects on glomerular filtration, and therefore on the filtration of sodium and calcium. It lowered urine calcium – as expected – but did not change urine sodium excretion.
Fasting (circles) and fed (triangles) but not overnight (squares) the drug lowered urine calcium excretion (Figure to the left, upper left panel) shown here as millimoles/hour. The effect was statistically significant fed because of the large number of observations; fasting it was more marked but we had fewer measurements so formally speaking it was not significant. Overnight there was no change at all.
The fraction of the filtered calcium excreted (upper right panel) represents how the kidney tubule cells alter their reabsorption of filtered calcium – a highly regulated process. Both fasting and fed, that fraction fell markedly with chlorthalidone and both effects were statistically significant. Note that you can make these fractional excretions into percents – multiply by 100 (0.04 = 4%).
Lithium is everywhere, in our water and our food in minute concentrations. It has an odd property. Part of the kidney unit – the nephron -, that part closest to the glomeruli and therefore called the proximal tubule, handles lithium and sodium more or less equally, whereas the latter parts of the nephron do not handle them equally but let most lithium go by into the urine.
This tiny fact lets us sort out where along the nephron thiazides might have their main effects. Fasting and fed, chlorthalidone raised the fraction of filtered lithium reabsorbed in the proximal parts of the nephron so that less went forward into the later parts of the nephron (Lower left panel of the figure).
Here is a fact. Calcium and sodium and reabsorbed along the proximal nephron more or less in parallel. So by knowing the fraction of filtered lithium and therefore the fraction of filtered sodium sent forward, we also know the fraction of filtered calcium (lower right panel), That fraction falls with chlorthalidone. As a result, the early portions of the nephron conserve more calcium with the drug than without, and that is one part of why the urine calcium falls.
My article on idiopathic hypercalciuria is an excellent primer and overview of lithium clearances and proximal vs. distal nephron calcium handling.
Acidity of the Urine
Chlorthalidone, and presumably other thiazide type drugs make urine more acidic. Fasting (upper left panel of figure, circles) the effect is small: Chlorthalidone points (black circles) are just a little to the left – lower pH more acid – than without the drug. But overnight, the four black squares are far to the left meaning the urine is a lot more acidic with the drug. Fed (Upper middle panel) the same.
This matters a lot for supersaturation with respect to calcium phosphate, not so much for calcium oxalate. Therefore, although the calcium oxalate supersaturations with and without the drug are barely different (right upper and lower panels) those for calcium phosphate are much lower with the drug (lower left – fasting and overnight and lower middle panel – fed).
This makes thiazide type drugs ideal for preventing calcium phosphate type stones and calcium oxalate stones. They can prevent the former because they lower calcium phosphate supersaturation via both reduction of urine calcium excretion and urine pH. Calcium phosphate stones form on plugs of calcium phosphate in kidney tubules or perhaps just in the urine itself, driven by supersaturation.
Calcium oxalate stones are more complex. They form on the surfaces of the renal papillae over deposits of calcium phosphate in the kidney tissue (plaque). A crucial initial step in this process involves the laying down a film of calcium phosphate over plaque. Calcium oxalate then deposits from urine on top of this initial film to make the stone. So thiazide acts in two ways to reduce calcium oxalate stones: by lowering calcium phosphate supersaturation and therefore hampering formation of the calcium phosphate film needed for overgrowth of calcium oxalate on plaque, and by reducing calcium oxalate supersaturation itself.
Note that the nine trials concerned calcium stones, and did not always distinguish between those that did and those that did not contain appreciable percentages of calcium phosphate.
I have not discussed here how thiazide lowers urine pH. That would take us too far afield. The original article points to changes in intestinal uptake of alkali, which seems to be reduced by the drug.
In our paper reviewed above we found no changes in urine oxalate with chlorthalidone, and did not therefore include oxalate data in the tables. In the past we have published a rather large group of patients who were put on thiazide and exhibited no increase whatever in urine oxalate. I do not believe it is likely that thiazide treatment raises urine oxalate.
How Do Thiazides Lower Urine Calcium?
I have said it is not just by acting on the kidneys, for that is to close down the drain – the bathtub will overflow perhaps but you cannot reduce the flow out of the drain for long that way. This is a sometimes overlooked point when people speak about urine calcium losses.
What happened in our patients must have been a fall in absorption of calcium from food, or an increased uptake of calcium into bone, because multiple measurements of serum calcium – the water in the bathtub – were almost unchanged despite a fall in urine calcium.
But almost is not quite unchanged.
Serum calcium – the upper left panel of the figure, rose significantly with thiazide. The amount of calcium filtered from blood into the kidney tubules (lower left panel) did not change significantly – the error bars overlap -, because filtration itself – (upper right panel) also varied. As I already showed you, urine calcium fell (lower right panel).
So the picture does have in it a bit of the bathtub with a somewhat closed drain – the water level, serum calcium, rose.
Now, I am about to leap into conjecture: Possibly, this increase in serum calcium could promote near instantaneous movement of calcium into bone by sheer physical chemistry.
A body of work that is perhaps more obscure than it should be, edges toward the idea that a portion of bone mineral, brushite in fact, may be in physical chemical equilibrium with blood so that large amounts of calcium can move in and out of bone mineral without any necessary cellular control.
This latter is critical, because bone cell process hours to change in response to hormones whereas urine calcium, as an example, can rise within a few minutes and yet serum calcium remain unchanged or even rise slightly as seen here.
This colorful picture (shown below) is similar to much of what we have often discussed on this site. To the left is diet calcium coming in and leaving, the absorbed calcium entering the ECF or extracellular fluid – we have approximated this as blood. Kidneys can release calcium, as noted.
The big addition is that a large amount of calcium may well be circulating between blood and bone – the 6,000 mg in the middle of the picture. This exchange is affected by serum phosphate, drugs like anticoagulants that alter bone mineral modifiers, and other factors. Of the 6,000 mg of calcium that enter and leave bone every day, only 400 mg go into the slowly exchangeable bone pool which is our familiar hydroxyapatite – the majority of calcium phosphate stones and the principle stiffener of bone.
What if, when kidney conservation falls – the drain opens – calcium leaves the exchangeable pool so the ‘faucet’ opens. What if, when thiazides raise conservation, the slight rise of serum calcium – from diet inflow perhaps – is just enough to nudge calcium flow into bone?
In other words, envision that bathtub. As we eat, the drain opens, the diet ‘tap’ opens, and bone comes into a kind of balancing act: If diet inflow is large enough calcium flows into the exchangeable bone mineral – brushite – pool. If not, calcium flows out. Thiazide clearly tightens the drain, the diet is the same with and without the drug, so possibly diet calcium inflow is raising blood calcium a bit and calcium is moving into bone.
If I were young I would test this idea in humans.
Effect of Thiazide on Bone Mineral Balance
All this raises an obvious question: Do thiazides improve bone health, reduce fractures, increase bone mineral content?
A recent review summarizes a large amount of evidence that they do indeed reduce fractures. The data, from the Danish population, suggest that duration of continuous treatment is more important than dosage in fracture prevention. In particular, fracture risk appeared to increase during the first year of use but then fall progressively. A Cochrane analysis concludes that all evidence to date – 2011 – supported this idea of reduced fractures, albeit there were no prospective controlled trials. Several such trials subsequently confirmed that thiazides maintain or increase bone mineral. Dalbeth and colleagues describe a link between rise of uric acid and bone mineral with thiazide implying a possible causal linkage not otherwise tested.
The overall impression is that one would not win by betting against thiazide as helpful to bone.
By now you might be asking if stone formers are at risk for bone disease. A recent long term followup of nurse and physician cohorts showed an increased risk of wrist but not hip fractures. Our article on idiopathic hypercalciuria summarizes additional information linking it and stone disease to fractures.
How About Salt?
As their prime purpose diuretics cause renal sodium wasting so that for any given intake of sodium the total body sodium stores will be less than without the drug. This is perhaps a main reason why they can reduce urine calcium via increase of proximal tubule calcium retention and also reduce blood pressure.
From this, one might think that low sodium diet might be like thiazide and help with bone mineral balance. In the one really ideal study testing this idea it was correct, and I have summarized that work in another article on this site.
Clinically, reduced salt intake is critical for successful thiazide use. From its basic physiology, high sodium intake will essentially undo the effect of the drug, and raise urine calcium. More, thiazides cause losses of potassium and lower serum potassium, and this is worsened by high salt intakes.
A caution. Thiazides can lower serum sodium levels, especially in older people, and this site advises reduced sodium intake which could increase that risk. In general diagnosis of low serum sodium was made 19 days after initiation of treatment (95% CI 8, 31 days). For this reason physicians invariably monitor serum electrolytes after a few weeks of treatment, and should do so after a few months, and at reasonable intervals, as well.
Is One Drug Better Than Another?
I think so. For example, hydrochlorothiazide has been used 2 times a day in the stone trials, whereas indapamide and chlorthalidone are long acting and needed once a day. People can forget the second dose. It is said that hydrochlorothiazide causes less of a fall in serum potassium but at least in hypertension trials it is less potent so you need more and when more is used it is the same as the other two drugs with respect to potassium, and dosed twice daily.
I like to start with lower doses than used in the trials. For chlorthalidone, 12.5 mg (1/2 of a standard 25 mg pill) is enough for most people. For hydrochlorothiazide I like the 12.5 mg pill twice a day – short acting. For indapamide I like 1.25 mg pills. Chlorthalidone and indapamide are long acting so the pill is once a day.
What Have We Learned?
Thiazide drugs can reduce stone recurrence at least in part by reducing urine calcium loss and supersaturation. They act on the kidney but also seem to improve bone mineral balance and reduce fractures. While stone prevention is certainly not accomplished by one pill, thiazide is an important part of what physicians can offer for prevention of stones.
188 Responses to “THIAZIDE DIURETICS FOR STONE PREVENTION”
Do you have any thoughts on vitamin k2 for its reputed effects on bone health and its possible relevance to the calcium metabolism of stone formers?
Thank you so much.
Fredric Coe, MD
Hi Richard, A PubMed search ‘vitamin K AND kidney stones’ restricted to humans, turned up 16 articles. As rule the paradigm was that vitamin K dependent proteins containing gamma carboxy glutamic acid (GLA) inhibit crystallization so low vitamin K increases and high vitamin K decreases stones. The best clinical test of this idea was essentially negative but now enough to falsify the entire idea. No data support supplemental vitamin K as a prevention (a trial, for example). Good question, good idea, unproved. Regards, Fred Coe
Harvey Feldman, MD
Dr. Coe: I would like your opinion on the recent clinical trial published March 2nd in the new England Journal of Medicine that reported no benefit on kidney stone prevention from HCTZ in doses ranging from 12.5 to 50 mg/day. How do you explain these results that contradict previous studies?
Fredric Coe, MD
Hi Dr Feldman, Of no significance. There is already too much thiazide trial data for this one to change practice. This one used a short acting drug none of us in the business rely on, and I suspect it was their main problem. As well stone counting was complex. I plan a serious review of the trial perhaps as my next article. So take this as my best right now. Regards, Fred
Hello Dr. Coe, What are your thoughts on using a thiazide diuretic in pts with unknown stone type who refuse 24 hour urine collections.
Fredric Coe, MD
Hi AG, I take a very dim view of this. By contrast the kidney stone diet matches the ideal US diet and I can recommend it to all. Regards, Fred Coe