My friends in the humanities are more popular than I am. At parties they can talk about poems, plays, novels, the modern and the ancient all with authority and considerable charm. Of course every doctor gets a share of medical questions. But when that brief flurry ends, and people ask what I do, a kind of pall falls over the table, or the little group at a cocktail party. What can I say besides that I prevent kidney stones? After a few curious people ask about how I do this, and others mention advice everyone seems to have heard somewhere, my moment passes. Even worse, if pushed for details, I am up against the unmentionable business of sampling body fluids.
Which is exactly why I am writing this article. The 24 hour urine collection is at the very center of stone prevention as practiced. Likewise for supersaturation. To me both belong at that center. I have come to say why I think so.
The article is long but if you read it through there is a special extra value waiting at the end. Those of you who have signed up for my emails might want to read what I have written there before embarking.
The Supersaturation Hypothesis
This site is itself saturated with articles about supersaturation. It is all crystals can recognize as a driving force for their formation and growth. Physicians can and do use it as a guide to treatment. Patients can understand it because it has a powerful intuitive character and can be replicated in common household experiments.
It is true that the natural power of supersaturation over crystallization is diffused and complicated by myriads of molecules in urine and probably kidney tissue itself, molecules which delay crystal formation and growth and in fact create the ability of urine to maintain a supersaturation over days or even weeks which would not remain in any common laboratory solution that lacked these molecules.
No doubt the mass of crystal inhibitors is why although supersaturation can reasonably correlate with the types of crystals in stones normal people commonly produce urine of comparable supersaturation to that from stone formers. For this very reason, we have proposed that the urine supersaturation of an active stone former, whatever it may be, is too high and needs to be lowered.
Which brings us to the 24 hour urine: In what does one measure supersaturation, and why? Obviously it needs to be measured in urine, but why a 24 hour collection. What needs to be measured in that urine? Can we use the measurements from which supersaturation is calculated for other clinical purposes? Can these measurements inform patients about their lives and their successes or failures in implementing stone prevention?
Uric Acid Stones
No one doubts the direct immediate and overwhelming power of supersaturation to produce uric acid stones. Nor does anyone doubt the dependence of uric acid supersaturation on low urine pH, so much so that a trial of potassium citrate – which raises urine pH – for uric acid stones will never be done: The effects are too obvious to warrant a trial.
I cannot but linger a moment, here, in this garden of certainty. For uric acid all is as it should be. Physics dictates rational treatment which works so well no one quibbles, no one doubts for a moment what is the proper course of action.
The other organic crystal based stone forms when urine is supersaturated with this poorly soluble amino acid because transport defects in the kidneys lead to excessive excretion. Though not easy to measure urine cystine supersaturation is a commercially available product that gives reliable results even in the presence of thiol binding drugs, and is widely used in patient management.
Unlike the case of uric acid stones, there is no magic potion to abolish cystine supersaturation. As a result treatment outcomes are not so sparkling as for uric acid. But even so, no one doubts for a moment that simple supersaturation causes cystine crystals and cystine stones, and that reduction of supersaturation is the key to their prevention.
The Vexed Topic
Supersaturation and the calcium stones remains a vexed topic simply because the linkages are less obvious in real life. Perhaps it is indeed the urine inhibitors which make it so. But I vote for supersaturation on grounds theoretical and by analogy: In the two common enough instances where things are clear it does what one expects.
The Components of Supersaturation Calculations
How to Calculate Supersaturations
You might say, why calculate supersaturation; why not measure it directly?
We did measure it directly, long ago, but the measurements are tiresome and expensive, and therefore impractical for large scale use for the millions of people who form stones.
To calculate it we use updated versions of a program written originally by Birdwell Finlayson whose picture is featured in this article. I knew him well, respected him as a brilliant scientist and supportive colleague, and felt his passing as a considerable personal loss.
What Measurements are Needed?
His program uses 13 measurements, and these have become the standard 24 hour urine ‘panel’ which is widely available from commercial vendors at a reasonable price.
On the one hand they are combined into the three supersaturations which correspond with the the three common stone crystals: calcium oxalate, calcium phosphate, and uric acid. On the other, each of them has other uses in clinical practice and it is these I mean to emphasize because altering them is what we do when we try to prevent stones.
The components needed for supersaturation are (all but pH are amounts in 24 hours): Volume, calcium, oxalate, phosphate, citrate, magnesium, pH, uric acid, sodium, potassium, chloride, sulfate, and ammonium ion.
Urea is usually included because it gives a measure of daily protein intake. Creatinine is included because it gives an estimate of completeness and consistency of collections. Neither are used to calculate supersaturations.
I focus here on what I call key measurements: Those which most strongly affect supersaturation and predict stone risk.
The others, mainly reflections of diet, are of immense importance in practice as they affect the key measurements and their variation via counseling is a principle component of medical stone prevention. I plan to discuss them in another article, soon.
Key Measurements Vary Over 24 Hours
Which Measurements are Key?
For calcium stones, volume, calcium, oxalate, citrate, phosphate, and pH are the most central measurements as they have the most effect on supersaturation and have been most clearly associated with stone risk.
For uric acid stones, it is volume, urine pH, and total uric acid excretion.
Why is Diurnal Variation important?
Peaks and Valleys are Averaged Out
The 24 hour urine, important as it is, reflects a compromise between what is true and what is practical. Such collections average out the excretion rates and concentrations of all the measurements and therefore the supersaturations, but can be obtained with a reasonable amount of effort from patients and produced commercially at acceptable prices. Lost are the peaks and valleys, irreparably so.
All I can do here is show what is lost, so everyone who collects and interprets these basic collections upon which rest evaluation and treatment of kidney stones understands what extremes are hidden in the polite averages we can obtain in practice.
Certain Periods are Most Perilous
I will show that nighttime is the most likely period when crystals will form, the fasting period just after arising being next. Physicians and their patients might want to consider this in planning treatment, especially the timing of fluids. This linked article alludes to and partly illustrates the time of day variations of supersaturations but does not give any information about the individual urine components responsible.
Spot Urines are Not Practical
On the other hand, the 24 hour collection is necessary just because of the marked variation of excretion rates, concentrations, and supersaturations during the day and night. If one wanted to avoid the annoyances of collecting a full 24 hour urine collection, when would one sample within the day? I will show that there is no realistic time. Overnight is dramatic concerning supersaturation, but misses the key excretion rates which we actually use for diagnosis and treatment.
The Diurnal Variations of the Key Measurements
Just because the issues of urine risk factors and supersaturations and their relation to calcium stone formation are a matter of ongoing debate, I have chosen our sets of data from calcium oxalate and calcium phosphate stone formers who have been studied in our clinical research center. All are eating the same diet during the measurements, and the measurements extend throughout a full day and overnight. Reasonably well matched controls are included as a contrast.
In these same patients, I have also included uric acid excretion and urine uric acid supersaturation, as a kind of alternative counterpoint. These patients do not produce uric acid stones, nor make stones which include any uric acid, so their uric acid supersaturations are expected to be about the same as in normal people.
Caution and Reservation
A note of caution and reservation was sounded by Dr. John Asplin who read the preliminary version.
These people were living in our clinical research unit. They drank as they chose but were collecting urine samples during the day. So no doubt they drank fluently and perhaps more than usual.
Someone doing orthopedic surgery, driving an 18 wheel Semi, teaching nursery school classes, or putting up house framing in the summer might have as low a urine volume by day as by night. So all I mean by the figures is the general sense that overnight and fasting are rigorously low volume, daytimes more liberal as a rule but with multitudes of exceptions.
A clinical corollary: Ask your patient about work, driving, airplanes; tell your doctor about splitting 5 cords of wood for your winter fuel.
All this refers mainly to urine volume but there is more that happens out on the street. Not everyone eats three meals a day. Some people snack or starve until supper. Many vary so much from day to day no urine sample will be ‘representative’.
Take what I show as a kind of principle or idealized view of things. Work from there, be you a patient or a physician.
The figures that follow have been re-envisioned and re-made by me for the purposes of exposition on this site. Published figures and tables from this set of data can be found in the references of these articles: Bergsland et al; Worcester et al.
Surely we have spoken enough of this humble and essential measurement. Physicians try to raise it. Patients need somehow to find ways of drinking enough and not go mad. Beverages themselves pose hazards and opportunities. Even the sceptical and faintly absurd guidelines released by the American College of Medicine support an increase of urine volume as a stone prevention. These guidelines are absurd because while recommending we increase urine volume they do not suggest we measure urine volume.
The truth about urine volume is that a 24 hour urine is at once invaluable and probably incomplete.
Look at the picture. The volume is shown in milliliters per hour; 30 ml is a shot glass – an ounce. The work was done under controlled diet conditions. Amounts of fluids were left to the patients’ choices.
B, L, and S are breakfast, lunch, and supper; H is home. ON is overnight. Red are normal people. Green and blue are calcium oxalate and calcium phosphate stone formers. The lines are error bars – standard errors for those who are interested.
We have already made much of the fact that supersaturations vary inversely with urine volume – higher volumes, lower supersaturations. It is true that we eat during the day, and it is true that what we eat comes into the urine not so long after we eat it, but it is not daytime we should most worry about. It is nighttime. The 24 hour urine is a wonderful test system, but it tends to underestimate the problem of high supersaturation.
This means that wise physicians and wise patients aim high – push urine volumes as high as possible. And, they are afraid of moonlight.
Apart from its being a critical part of stone crystals we have not made much of urine calcium up until now, because I wanted
a large and strong foundation about stones, inhibitors, supersaturation, and citrate before taking it on. But here we are.
The bars are like the first graph. The left graph shows loss rate into the urine; the right shows the concentration in millimoles – thousandths of a mole – per liter. I have explained molarity elsewhere; it is simply a common unit for expressing how many molecules or atoms are present in a unit of volume, like a liter. A mole of any molecule or atom contains exactly the same number: 6.023*1023 or 6.023 with 23 0’s after it – Avogadro’s number.
The left panel resembles the urine volume graph. Urine calcium rises with meals and falls overnight.
But – a huge difference – the amount lost in the urine is much higher in the two kinds of patients (blue and green) than in the normal people (red). This higher calcium loss has been noticed in calcium stone formers for more than 70 years, and it called ‘idiopathic hypercalciuria (IH)’ meaning high urine calcium of unknown cause.
Because of IH, the calcium concentrations among patients far exceed those of the normals, most especially overnight when their urine calcium levels remain higher than normal and their urine volumes fall even lower than normal. These concentrations are shown in the right panel, in mmol/liter.
Think about this: While stone formers with IH sleep, they conserve water as they should, but lose calcium into the urine. Concentrations, simply the ratio of one flow – calcium – to another – water, they rise to the highest levels of the 24 hour period.
Once again, the 24 hour urine, our gold standard and ideal testing vehicle, underestimates problems, and what I have already said about wise physicians and wise patients continues to hold.
Like calcium, citrate excretion rises promptly with meals, and stays up all day. But unlike calcium, citrate losses of patients are not higher than those of normals, but about the same. Because of higher urine volumes during the main part of the day, urine citrate concentration tends to fall in the afternoon, whereas IH maintains urine calcium concentration at high levels in patients. At night, citrate concentration rises, but as compared with calcium there is no contest. Calcium molarity is about 6, citrate less than 2 – remember, molarity matches the molecules in numbers per volume, even though they are of different sizes.
Because citrate is so much lower in concentration than calcium, especially overnight, much of the calcium is free to combine with oxalate or phosphate to form stones.
The other half of the common calcium oxalate stone, oxalate is present at a very low amount and concentration compared to calcium, but changes in its concentration alter supersaturation in exactly the same proportions as chances in calcium,
they being equal partners. A long standing and silly myth held that changes in oxalate concentration raised supersaturation more than equal – in molar terms – changes in calcium concentration, elevating oxalate to a higher platform. In fact like all myths this one was easily proven false.
Many have said that urine oxalate excretion by stone formers is higher than normal, and the left hand panel shows this seems true even when diets are fixed and in fact contain very little oxalate. The difference from normal is fasting and the first 2 meal periods; thereafter excretions are all similar.
The concentration pattern, which determines supersaturation looks like that of calcium in having a nighttime peak, but the whole day is barren – the rise of oxalate excretion is small compared to that of urine flow, exactly the opposite of the case for calcium, so high concentrations are present only fasting and overnight.
The concentration of oxalate in urine is very small compared to calcium, and therefore the excretion rate is in umol/hour meaning micromole or millionths of a mole instead of the thousandths of a mole for calcium and citrate.
If you are beginning to see some real pattern here concerning the nighttime, you are seeing right. It is a terrible time to be a stone former.
The pattern of excretion is not the same as oxalate or calcium or citrate, but rather a delayed rise during lunch and supper and a marked fall overnight. As a result the concentration of urine phosphate is high fasting and rises progressively throughout the day into the night.
Whereas IH affects both kinds of stone formers, urine phosphate of the calcium oxalate stone formers (yellow) is above that of the calcium phosphate stone formers in the afternoon and above normals as well.
We have encountered pH before but in the form of alkaline and acid not the actual measurement. It is a logarithm to the base 10, so a single unit of change is a 10 fold change in acidity or alkalinity. The importance of pH is for calcium phosphate stones. Phosphate can combine with calcium only when it has 2 negative charges to match the 2 positive charges on the calcium atom. At pH of 6.8 1/2 of the phosphate in urine has the requisite 2 charges. At lower values of pH the fraction is much lower. For example at 5.8 it is less than 10%.
As you might expect, the urine pH of the calcium phosphate stone formers (blue) is higher than that of the calcium oxalate stone formers (yellow) and in fact higher than that of urine from normal people, too.
Overnight, calcium oxalate stone formers have a very low pH, so calcium phosphate crystals cannot form. Calcium phosphate stone formers have a much higher pH, so calcium phosphate crystals could form. Normal people are more like calcium phosphate stone formers than like calcium oxalate stone formers in urine pH, but lack IH so they do not have anywhere the same high levels of urine calcium molarity.
Urine Uric Acid
Concentrations are much like those of oxalate and phosphate with a distinct dip in the middle of the day and marked peaks overnight. There is little variation between the patient groups and between stone formers and normal people.
Note, these are calcium stone formers, so the uric acid excretions are not related to uric acid stones.
Like minor comics prepare an audience for the star, like a concert orchestra plays something by itself to accustom the audience to listening before the grand virtuoso seats herself before the keyboard, all of the main components of supersaturations have had their hour upon the stage.
As you might expect, calcium oxalate supersaturation is rather constant all day, but high at night. Likewise, that for calcium phosphate rises all day long and peaks at night.
But the dashed line at 1 tells a story. Normal people do not produce an average calcium phosphate supersaturation above 1 – the point where crystals can form – whereas both kinds of patients regularly do so.
Uric Acid Supersaturations
Being calcium stone formers whose stones contain no uric acid crystals so far as we could tell, one should expect no significant supersaturation for uric acid above that found in normal people who of course produce no stones at all.
In fact, only fasting average supersaturation rises above one (the dashed horizontal line), in normal people and calcium oxalate stone formers alike, and overnight – but slightly – in calcium oxalate stone formers. Presumably these values outline the permissible range of supersaturation that can be tolerated without stone formation.
The lack of impressive supersaturations, far below those for CaOx and CaP, arises entirely from the spectrum of urine pH values in these patients and normals. Uric acid stone formers have much lower urine pH values, and therefore higher supersaturations.
I have treated the main determinants of supersaturation with considerable respect and given to their analysis a large acreage on this site, yet have offered not a shred of proof that they or the supersaturations they engender cause stones. Even more, I have written this article about 24 hour urines and decomposed them into the meal periods and overnight as if the 24 hour urine itself were somehow less informative.
All this is fair if I have something special to offer, and as things turn out I do.
How Can We Test if Urine Determinants of Supersaturation Cause Stones?
Alter Them and Measure Stone Formation
Perhaps the perfect gold standard is to vary one of them at a time and measure new stone production. The trials done to date perhaps offer some evidence about this matter. Though I have not as yet reviewed them, three trials of thiazide have been positive, and that drug acts – so far as we know – mainly by lowering urine calcium excretion. Potassium citrate raises urine citrate and five citrate trials have proven positive. For water itself we have one reasonable trial which was positive. For oxalate and phosphate we have no trials, and for pH we have no trials that pertain to calcium stones.
Of course, for uric acid, we have no trials because the effects of raising urine pH are so dramatic no trial seems warranted.
Are they Associated with New Stone Onset?
This is an epidemiology question, and fortunately we have an excellent answer. Professor Gary Curhan has performed superb analyses of three large groups of people who have volunteered to offer health information to investigators over many years of their lives. Two are composed of women who are nurses; one is composed of men who are doctors.
In each cohort, over time, some people began to form kidney stones and had not done so in the past. Gary was able to obtain 24 hour urine samples on a significant sub-set from all three cohorts and could compare the excretion rates of the key urine constituents – calcium, oxalate, citrate, and volume as well as other measurements in people who became stone formers, and as a control in groups of people otherwise similar who had not become stone formers.
For each of the 6 ranges of urine calcium excretion on the horizontal axis of the upper left hand graph, Gary offered a mean relative risk ratio, along with the upper and lower 95% confidence interval for that mean. I have chosen the lower 95% border to plot as a bar: when that passes through 1, meaning risk is present, one can be very confident that is indeed the case, as I have chosen a very stringent requirement.
All three cohorts – the two female cohorts in red and the male cohort in blue show risk is certainly present above 200 mg/day of urine calcium. Values of this level are not rare among even normal people, and certainly are very common among stone formers with IH.
Of importance, the higher the urine calcium excretion, the higher the relative risk of being a stone former. This is very important. If something is a cause of stones, more of it should cause stones with more certainly.
For any random assortment of urine volumes, the higher the daily excretion of calcium the higher the calcium concentration, and, all things being equal, the higher the urine calcium oxalate and calcium phosphate supersaturations. So his associational findings support both the supersaturation hypothesis and the value of making 24 hour urine collections the basis for evaluation of patients.
The exact graphical format yields the same results for urine oxalate excretion – upper right panel – but the beginning of risk is very low, 25 mg of oxalate daily in either sex. I have not as yet provided 24 hour urine oxalate excretions on this site, but will tell you here that most patients are above 25 mg daily so risk is present much of the time. Even so, as urine oxalate rises risk increases pari passu meaning it is a reliable graded risk factor.
Citrate runs backward from calcium and oxalate – it is against stones because it can bind calcium and inhibit calcium crystal formation. Therefore I have plotted the upper 95th percentile, meaning that when it falls below 1 levels are high enough to reduce excess risk to 0. Uniform high risk begins only when urine citrate falls below 400 mg, in either sex. Like oxalate and calcium, the risk profile of citrate is graded by excretion rate, and risk falls progressively as urine citrate rises.
Just like citrate, higher volumes are protective so I have plotted the upper 95th percentile. Risk falls progressively as urine volume rises. Above about 2.25 liters daily the risk ratio is safely below 1. I have used this threshold in my analysis of how to prescribe fluids.
What Do We Conclude From What I Have Shown?
The idea that the drivers of supersaturation cause stones is an hypothesis, and the two tests I have presented are deductions from that hypothesis. Both turn out to be true in fact. This means the idea is workable for now until disproved by other tests.
It does not mean, however attractive the idea, that supersaturation is itself a driver of stones. That would require that supersaturation itself were the target of treatment trials, and of epidemiological association studies. Neither has been the case.
I could argue forcibly that supersaturation is indeed the correct statement of the driving force for crystal formation and growth, and no one could deny me as this is a physical law and there is only one physics.
But because of the complex molecular environment of urine supersaturation may not itself be as important as the kinetic retardants of crystallization in urine.
Likewise, because almost all of our measurements use 24 hour urines which blur the variations during the day, the test may lack sensitivity enough to even perform meaningful tests.
This means we need trials and epidemiology to test the specific point. I predict supersaturation will prove a ruling variable once it is tested properly.
Let’s Have a Practical Summary
Why not collect morning or afternoon spot urines and save the laborious process of a 24 hour urine collection? The answer is in the graphs. When, exactly will you want to collect? The price of the measurements is the same, almost, for a spot as for an entire 24 hour collection. Many spots during a day will be expensive and also tedious. I think the idea is not sound and the graphs say so.
The worst supersaturations occur at night, so why not just collect for the 8 hours from late evening to morning?
For supersaturation maxima this is not a bad idea. But excretion rates of patients and normals become very similar overnight, so what you can manipulate – excretion rates – are harder to quantify. Look at citrate for example. More serious, the exquisite data linking excretion rates to stone risk will be useless, as they are based on 24 hour collections.
Do Without and Play the Odds
It is true that stone formers have lower volumes, higher excretions of calcium and oxalate, and lower excretions of citrate than normal people, but any one person may have only one of these abnormalities, or none, or many. How can you treat someone for everything if you know nothing? What will patients think who are offered blind treatment across all modalities? Why will a patient adhere to treatments based on no specific knowledge of his or her biology? Even more, if you do treat, how will you or your patient know if anything happened?
Get 24 hour Urines to Plan Treatment
That is my message. I say it is idiotic to do otherwise, and a vast disservice to patients who come for stone prevention.
How Many 24 hour Urines?
In my own work I measure 3 baseline 24 hour urines before beginning treatment. A single large practice collaborator collected two such urines before treatment. This gave us an opportunity to compare the first and second of each group from the practice and the first vs. the third from my own practice asking how much variation is present.
The figure shows data for only urine calcium, oxalate, and citrate; men and women are shown in the upper and lower panels, respectively.
In each individual panel, data from my practice is outlined by a solid ellipse, data from the collaborator by a dashed ellipse. The ellipses are calculated to contain 68% of the data points on each panel.
It is obvious that values from the two urines are highly correlated with one another. The ellipses are narrow and point upward and to the right at about 45 degrees. This fact has mislead other scientists into making the wrong recommendations. Charles Pak, whose work I admire said that given the correlation a single urine is sufficient. Likewise doctors Eisner and Stoller have come to the same conclusions.
But I believe they are wrong.
If one wants to predict the average values for a group of patients, the strong correlations make it likely that one urine will do. But – that word! – for any one individual person the value from one urine can be very far from the value in the second urine.
To see this, simply draw or imagine a vertical line at 200 mg of calcium in the lower left panel of the figure. It intersects with the bottom of the ellipse at about 120 mg and with the top of the ellipse at about 300 mg/24 hour. Given Gary Curhan’s work which I have summarized in the prior section, the first urine could place a patient at no risk for stones from urine calcium excretion, the second place the same person as high risk. Since the ellipse contains only 68% of the data, the actual spread can be much larger.
You can see the same by picking points. It is easy to find points at 150 mg/day along the x axis and 300 mg/day on the y axis.
You Need at Least Two
Given the high probability of large errors in diagnosis and therefore treatment, at least two urines are required. Moreover, their duplicative powers can be helpful. For example, if someone has a urine calcium excretion of 300 mg one day and 150 the other, that person has some factors in life that can regulate urine calcium from high risk – by the Curhan criteria – to low risk and those factors, of diet or lifestyle, can be used as an aspect of treatment. How to do this is for other posts, as it is a complex matter, but the option is obvious and important.
I hope Drs Pak, Eisner, and Stoller do not resent my concluding other than they did, and I invite their rebuttal and argument. But to me their data look much the same as mine, and we differ mainly in how to interpret it.
How You Choose Determines What You Get
What I have not said is perhaps too obvious: A 24 hour urine is one frame out of a movie that is life long. How you pick the right days, and see to it they represent a decent sample of life as lived, that special choice determines how useful the samples can be. Physicians can be aware of the importance of such choosing, but it is only patients who can make the choices properly.
Where else could I have placed this crucial message better than at the end?
At the beginning it would be lost in the general mess of information. Here it is a last word.
Admittedly, not everyone will have the patience to finish this very long article but for those who do I offer this mote of potential value and reward.