# The Metabolic Stone Workup & the Evidence Behind Stone Prevention

## Executive Summary

**Key Points**

- Recurrence prevention rests on a coherent physiologic model (lower the supersaturation of the offending salt) but a surprisingly thin and aging randomized-trial base. The widespread instinct that much of practice is closer to eminence than evidence is largely correct.
- The two largest, most rigorous modern RCTs in common stone disease both returned **null** primary results: NOSTONE (hydrochlorothiazide, n=416)^([4](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-4)) and PUSH (behavioral hydration, n=1,658)^([1](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-1)).
- The classic positive trials^([2](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-2),[3](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-3),[6](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-6),[7](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-7),[8](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-8),[9](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-9)) are small, mostly single-center, decades old, and several inflate apparent efficacy via historical/pre-treatment control rates — the documented “stone clinic effect”^([13](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-13)).
- **On the added questions:** oral calcium with meals lowers urinary oxalate but has no hard-outcome RCT; **magnesium** supplementation has essentially **no** supportive RCT (the Mg(OH)₂ arm of Ettinger 1988 was not superior to placebo)^([6](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-6)); **citrate in calcium-phosphate** disease is extrapolated and physiologically double-edged; and **allopurinol’s** randomized evidence is for hyperuricosuric calcium-oxalate stones^([7](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-7)), **not** uric-acid stones.
- The only crisp, modern, randomized oxalate data come from the orphan space: **lumasiran** cut urinary oxalate approximately 65% vs placebo in primary hyperoxaluria type 1 — on a surrogate, not stone events^([14](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-14)).
- **By endpoint:** one old trial shows fluid reduces symptomatic events^([2](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-2)); modern trials capturing radiographic new-stone formation and growth of existing stones were null^([1](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-1),[4](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-4)); the only positive randomized signal on stone **growth** is acetohydroxamic acid in struvite^([12](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-12)).

## 1. The Central Tension: Plausible Physiology, Thin Trials

Kidney-stone prevention is built on an elegant and almost certainly correct physical-chemistry premise: a stone forms when urine is supersaturated with respect to a crystalline salt, and lowering the relative supersaturation of that salt should reduce crystallization. Every standard intervention maps cleanly onto this model — dilute the urine, lower the cation, lower the anion or its precursor, raise an inhibitor, or shift pH out of the salt’s crystallization window.

The difficulty is that physiologic plausibility and supersaturation reduction are surrogate arguments, and the randomized evidence that these maneuvers reduce clinical stone events is far weaker than guidelines imply^([10](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-10),[11](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-11)). The literature is dominated by small, single-center trials from the 1980s–early 2000s, many comparing on-treatment recurrence to a pre-treatment historical rate — a design biased toward apparent benefit. When the field finally produced large, modern, placebo-controlled trials, the two flagship results were negative^([1](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-1),[4](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-4)).

## 2. The Metabolic Workup: Who, What, and When

Stone analysis is the single highest-yield test and should be obtained whenever a stone is retrieved or passed; it reorganizes the differential and frequently overrides the urine chemistry. A basic evaluation (serum calcium, electrolytes/bicarbonate, creatinine, urinalysis with pH; plus stone composition and imaging) suits any stone former. The comprehensive evaluation — two 24-hour urine collections — is reserved by guideline for recurrent, high-risk, or interested first-time formers^([11](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-11)), since most single-stone formers will not be shown to benefit from drug therapy.

Two non-consecutive 24-hour collections on the habitual diet are standard, because day-to-day variability is large. Defer collection several weeks after an acute episode or procedure. Serum studies screen for primary hyperparathyroidism, distal renal tubular acidosis (non-anion-gap acidosis with inappropriately alkaline urine), and CKD.

Clinical Pearl

Stone composition trumps urine chemistry. Hypercalciuria means something different if the stone is calcium oxalate (treat the calcium) versus calcium phosphate at pH 6.8 (now you worry about distal RTA and you are wary of alkali). Always anchor on the stone first.

## 3. Reading the 24-Hour Urine (Litholink-Style Report)

Commercial panels report raw 24-hour excretions, volume and pH, dietary surrogates (sodium, sulfate, urea nitrogen, protein catabolic rate), and — most usefully — calculated relative supersaturations (SS) for calcium oxalate, calcium phosphate (brushite), and uric acid. Read in three passes: (1) check collection adequacy, (2) read the supersaturations to identify the threat salt, then (3) read the analytes to explain why SS is high and what is modifiable.

**Collection adequacy first.** Confirm completeness using 24-hour creatinine indexed to body weight (approximately 15–20 mg/kg/day men, approximately 10–15 mg/kg/day women). An implausibly low creatinine signals under-collection that spuriously lowers every analyte. Misreading an incomplete collection as “normal calcium” is among the commonest interpretive errors.

**Conventional analyte targets** (physiologic targets reflected in the AUA framework^([11](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-11)), not thresholds validated to predict prevention — treat as directional):

| Parameter | Conventional target / abnormal threshold | Interpretive note |
|----|----|----|
| Volume | \>2.0–2.5 L/day (target ≥2.5) | Universal lever; lowers SS of every salt |
| Calcium | Hypercalciuria \>250 mg (W) / \>300 mg (M), or \>4 mg/kg | Commonest abnormality in Ca stones |
| Oxalate | Hyperoxaluria \>40–45 mg/day | Diet/gut driven; enteric or primary if very high |
| Citrate | Hypocitraturia \<320 mg/day | Inhibitor; low with acidosis, hypokalemia, high protein, dRTA |
| Uric acid | Hyperuricosuria \>700 mg (W) / \>800 mg (M) | Promotes CaOx and is the solute for UA stones |
| Urine pH | Interpret by stone type | Low (\<5.5) → UA; high (\>6.5–7) → CaP / struvite / dRTA |
| Sodium | Surrogate for intake; aim \<100–150 mEq/day | Drives calciuria; lever behind “lower salt to lower calcium” |
| Sulfate / urea / PCR | Animal-protein intake surrogate | High protein raises Ca and UA, lowers citrate and pH |
| Supersaturation (CaOx, CaP, UA) | Goal: lower toward / below 1 | The integrated readout; the target and adherence tracker |

Clinical Pearl

Use supersaturation as the dashboard. Analytes tell you which knobs to turn; the SS for the offending salt tells you whether you turned them enough. Falling SS is the most concrete evidence a regimen is doing what it mechanistically should — though SS reduction is itself a surrogate (Section 9).

Warning

The 24-hour urine has never been shown in a randomized trial to improve stone outcomes *as a test* (panel-guided vs empiric therapy). Its justification is mechanistic and pragmatic, not outcome-validated.

## 4. Stone Types: Risk Factors and Directed Therapy

### 4.1 Calcium oxalate (approximately 70–80% of stones)

The workhorse stone and the one nearly all trial evidence actually addresses. Risk factors: low volume, hypercalciuria, hyperoxaluria, hypocitraturia, hyperuricosuria; urine pH is largely irrelevant to CaOx crystallization. Directed therapy: fluid for all; thiazide for hypercalciuria; potassium citrate for hypocitraturia; dietary sodium and animal-protein restriction with maintained normal calcium; oxalate moderation with adequate calcium; allopurinol when hyperuricosuria is present.

Clinical Pearl

Counsel normal — not low — calcium intake. Dietary calcium binds oxalate in the gut; restricting it raises urinary oxalate and paradoxically increases CaOx risk^([3](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-3)). A frequent patient error.

### 4.2 Lowering urinary oxalate: diet, calcium binders, B6, and RNAi

Oxalate can be attacked at intake, gut absorption, and hepatic production — interventions that differ enormously in evidence quality.

- **Dietary oxalate restriction + maintained calcium.** Lowering high-oxalate foods while keeping normal calcium intake (so gut calcium binds oxalate) is first-line. Borghi 2002 is the supporting signal — oxalate fell on the normal-calcium arm and rose on the low-calcium arm^([3](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-3)).
- **Oral calcium *with meals* as an oxalate binder.** Calcium supplements taken with meals (calcium **citrate** is often preferred because it also supplies citrate; calcium carbonate or **calcium acetate** also bind oxalate) reduce oxalate absorption — the rationale in enteric hyperoxaluria (bariatric surgery, IBD, short gut, fat malabsorption). Evidence is physiologic and observational; small studies show lower urinary oxalate, but there is **no randomized hard-outcome (stone) trial**. Timing is everything: calcium *with* meals binds oxalate (helpful); calcium *between* meals raises calciuria (harmful).
- **Bile-acid sequestrants (cholestyramine)** are used in enteric hyperoxaluria to bind bile salts/fatty acids and indirectly reduce oxalate absorption — observational only.
- **Pyridoxine (vitamin B6)** lowers oxalate in a subset of primary hyperoxaluria type 1 (B6-responsive AGXT genotypes); observational, not an idiopathic-stone therapy.
- **RNA-interference agents (lumasiran, nedosiran).** For primary hyperoxaluria, lumasiran reduced 24-hour urinary oxalate by approximately 65% versus placebo (least-squares mean difference −53.5 percentage points; 84% reaching ≤1.5× ULN vs 0%) in the phase-3 ILLUMINATE-A trial^([14](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-14)); nedosiran is a same-class agent. These are the only methodologically strong, modern, randomized oxalate data — but in a rare genetic disease and on a urinary-oxalate **surrogate**, not stone events.

Clinical Pearl

The evidence gradient for oxalate lowering is steep and inverted relative to use: diet/binders (used constantly, observational) → B6 (niche) → RNAi (rare disease, rigorous RCT, surrogate endpoint). The most rigorous oxalate evidence applies to the fewest patients.

Warning

Calcium-based oxalate binders for common idiopathic CaOx stones rest on physiology and small studies — no randomized trial shows they prevent stone events. Counsel timing-with-meals and avoid implying proven outcome benefit.

### 4.3 Calcium phosphate (rising, especially in women) — and citrate’s double edge

Driven by higher urine pH (favoring brushite/apatite), hypercalciuria, and hypocitraturia. CaP stones should prompt evaluation for primary hyperparathyroidism, distal RTA, and sometimes medullary sponge kidney. Almost none of the randomized treatment evidence was generated in CaP formers — the trials enrolled CaOx disease — so management is largely extrapolation.

Citrate in CaP disease is the clearest example. The citrate RCTs^([8](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-8),[9](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-9)) enrolled calcium (predominantly oxalate) stone formers, not a CaP cohort. Mechanistically citrate is double-edged: it corrects hypocitraturia and chelates calcium (good), but the alkali load raises urine pH, which **increases** calcium-phosphate supersaturation (bad). Net effect in CaP disease is genuinely uncertain, and there is no dedicated randomized trial to settle it.

Warning

Adding potassium citrate to a calcium-phosphate former can be counterproductive: raising urine pH may increase CaP supersaturation even as it corrects hypocitraturia. If used, monitor follow-up pH and CaP SS rather than assuming the CaOx benefit transfers.

### 4.4 Uric acid (approximately 8–10%)

The defining driver is persistently low urine pH (\<5.5), usually from impaired renal ammoniagenesis tied to insulin resistance, obesity, and type 2 diabetes; low volume and hyperuricosuria contribute. Uric acid solubility is exquisitely pH-dependent, so the cornerstone is urinary alkalinization to pH approximately 6.5–7.0 (potassium citrate), which can both prevent and chemically **dissolve** existing uric-acid stones; fluid and metabolic/weight management support this. Allopurinol is reserved for hyperuricosuric or refractory cases (see 6.1).

Warning

Despite being the most mechanistically certain of all stone therapies, urinary alkalinization for uric-acid stones has essentially **no** randomized controlled trial evidence for clinical event reduction. It is standard of care on physiologic and dissolution grounds.

### 4.5 Struvite / infection stones (magnesium-ammonium-phosphate)

A microbiologic problem, not a metabolic one: urease-producing organisms (Proteus, Klebsiella, others) hydrolyze urea, alkalinize urine, and precipitate struvite, often as staghorn calculi. Primary therapy is complete surgical clearance (percutaneous nephrolithotomy), because residual fragments harbor organisms and seed regrowth; targeted antibiotics support eradication. The urease inhibitor acetohydroxamic acid (AHA) is the only agent with randomized evidence^([12](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-12)) but is poorly tolerated and adjunctive/palliative.

Clinical Pearl

For struvite, the metabolic panel is secondary. Source control — complete surgical removal plus organism eradication — is the treatment, and it became standard of care without a randomized trial because the pathophysiology and surgical necessity are unambiguous.

### 4.6 Cystine (rare, genetic)

Cystinuria causes high cystine excretion and recurrent stones from childhood. Management: high-volume fluid (often 3–4 L/day), aggressive alkalinization (pH \>7–7.5), sodium restriction, and thiol-binding agents (tiopronin, penicillamine) for refractory disease. Evidence is observational; randomized data are minimal given rarity.

## 5. Does It Actually Work? Evidence for Each Therapy

Each standard intervention paired with its best primary trial, the observed effect, and a candid grade. Pooled relative risks are from the AHRQ/ACP systematic review (28 RCTs)^([10](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-10)).

| Risk factor → therapy | Best primary trial(s) | Effect observed | Evidence grade |
|----|----|----|----|
| Low volume → fluid | Borghi 1996^([2](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-2)); PUSH 2026^([1](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-1)) | Borghi: 5-yr recurrence 12% vs 27% (p=0.008); PUSH: program NULL | Pooled RR 0.45, **low** strength; 1 small positive + 1 large null |
| Hypercalciuria → thiazide | Ettinger 1988^([6](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-6)); NOSTONE 2023^([4](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-4)) | Older positive; NOSTONE NULL at all doses, no dose-response | Pooled RR 0.52 (older) **contradicted** by best modern RCT |
| Hypocitraturia → K-citrate | Barcelo 1993^([8](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-8)); Ettinger 1997^([9](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-9)) | 1.2→0.1 stones/yr vs placebo unchanged; K-Mg citrate RR 0.16 | Pooled RR 0.25 (largest effect) but small, single-center, dated |
| Hyperuricosuria + CaOx → allopurinol | Ettinger 1986^([7](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-7)) | Fewer calculi vs placebo; only in hyperuricosuric/hyperuricemic | Pooled RR 0.59; confined to hyperuricosuric subgroup |
| High Na/animal protein → diet | Borghi 2002^([3](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-3)) | Normal-Ca/low-Na/low-protein vs low-Ca: RR 0.49 (0.24–0.98) | Single trial; benefit ran through **lower oxalate**, not lower Ca |
| High oxalate → diet + Ca binder | *(no stone-outcome RCT)* | Lowers urinary oxalate; binds gut oxalate with meals | **Physiologic / observational**; no randomized event data |
| Primary hyperoxaluria → lumasiran | Garrelfs 2021^([14](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-14)) | 24-h urine oxalate −65% vs placebo; 84% ≤1.5× ULN vs 0% | Strong RCT but **surrogate** endpoint, rare genetic disease |
| Putative inhibitor → magnesium | Ettinger 1988, Mg(OH)₂ arm^([6](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-6)) | Mg hydroxide **not** superior to placebo; chlorthalidone beat both | **Negative / no benefit** as monotherapy |
| Uric-acid stone → alkalinization | *(no RCT)* | Prevention/dissolution on physiologic grounds | **Expert opinion / mechanism**; no randomized event data |
| Uric-acid stone → allopurinol | *(no RCT for UA stones)* | Lowers uric acid; adjunct only, not first-line | **No** randomized evidence for UA-stone prevention |
| CaP stone → citrate | *(no dedicated CaP RCT)* | Corrects hypocitraturia but raises pH → may ↑CaP SS | **Extrapolated** from CaOx trials; double-edged |
| Struvite → surgery + AHA | Griffith 1991^([12](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-12)) | AHA: stone growth 17% vs 46% placebo; 22% intolerable AEs | Surgery standard w/o RCT; AHA positive but toxic, adjunctive |

Clinical Pearl

The strongest single number — citrate’s pooled RR 0.25 — comes from the trials with the weakest modern replication. The most rigorously tested calcium-stone intervention — thiazide in NOSTONE — is the one that failed. There is an inverse relationship between trial quality and apparent effect.

## 6. Disaggregating Three Commonly Conflated Therapies

### 6.1 Allopurinol: two different questions, one answer

Allopurinol is asked to do two unrelated jobs, and only one has randomized support. **For calcium-oxalate stones with hyperuricosuria**, Ettinger 1986 randomized 60 such patients and found fewer calculi on allopurinol^([7](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-7)); the AHRQ pooled RR is 0.59, with benefit confined to those with hyperuricemia/hyperuricosuria (mechanism: reducing monosodium-urate–induced salting-out of calcium oxalate). **For uric-acid stones**, allopurinol lowers urate production and is mechanistically attractive, but there is no randomized trial showing it prevents uric-acid-stone recurrence — alkalinization is the cornerstone, and allopurinol is reserved for hyperuricosuric or alkalinization-refractory disease. So allopurinol’s evidence base is a calcium-oxalate therapy that happens to act on urate, not a uric-acid-stone therapy.

### 6.2 Magnesium: a popular supplement essentially without a trial

Magnesium complexes oxalate in vitro and is widely sold as a “stone” supplement, but the randomized evidence is negative or absent. In Ettinger 1988, the magnesium-hydroxide arms (650 and 1,300 mg/day) showed reductions from predicted rates that were **not significantly better than placebo**, while chlorthalidone significantly outperformed both^([6](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-6)). The frequently cited positive trial^([9](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-9)) tested potassium-**magnesium** citrate, and its 85% risk reduction is attributed to citrate (and the combination), not magnesium; there is no quality RCT supporting magnesium monotherapy.

Clinical Pearl

When a patient asks about magnesium for stones, the honest answer is that the one randomized test of magnesium (as the hydroxide) did not beat placebo, and the positive “magnesium” trial was really a citrate trial. Magnesium is not an evidence-based anti-stone therapy on its own.

### 6.3 Citrate in calcium-phosphate disease (recap)

As in 4.3: citrate’s randomized evidence comes from calcium-oxalate cohorts; in calcium-phosphate disease the pH rise can increase CaP supersaturation, there is no dedicated RCT, and use is extrapolated and monitored rather than proven.

## 7. Preventing New Stones vs Growth of Existing Stones

Stone trials use heterogeneous endpoints, and lumping them overstates the evidence. Separate three outcomes: symptomatic stone events (passage or intervention), formation of **new** stones on imaging, and **growth** of pre-existing stones.

| Intervention | Symptomatic events | New radiographic stones | Growth of existing stones |
|----|----|----|----|
| Increased fluid | Borghi 1996 **positive**; PUSH 2026 NULL^([1](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-1),[2](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-2)) | PUSH: no difference | PUSH: no difference (≥2 mm) |
| Thiazide (HCTZ) | NOSTONE NULL^([4](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-4)) | NOSTONE NULL | NOSTONE NULL (enlargement) |
| Potassium citrate | ↓ stone-formation rate (older, small)^([8](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-8),[9](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-9)) | Positive in old trials | Not isolated as an endpoint |
| Allopurinol (CaOx + hyperuricosuria) | Ettinger 1986 **positive**^([7](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-7)) | Positive (new calculi) | Not isolated |
| Diet (Na/protein, normal Ca) | Borghi 2002 **positive**^([3](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-3)) | Implied by relapse reduction | Not isolated |
| AHA (struvite) | — | Less new growth^([12](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-12)) | **Positive: 17% vs 46%**^([12](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-12)) |
| Magnesium | Not superior to placebo^([6](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-6)) | Not superior | Not isolated |

Two honest conclusions follow. First, **the only positive randomized signal specifically on growth of existing stones is acetohydroxamic acid in struvite**^([12](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-12)) — for calcium stones, the trials that rigorously captured radiographic growth (NOSTONE, PUSH) were null^([1](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-1),[4](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-4)). Second, the older “positive” calcium-stone trials measured new-stone-formation rates against historical/pre-treatment baselines, the design most vulnerable to regression to the mean. The cleaner the endpoint and the more modern the trial, the weaker the apparent benefit becomes.

Warning

There is no convincing modern randomized evidence that any calcium-stone medical therapy halts growth of an existing stone. A stable stone on therapy is reassuring but is not, by itself, proof the therapy caused the stability.

## 8. Two Instructive Details the Guidelines Gloss Over

### 8.1 The famous diet trial did not work the way it is taught

Borghi 2002 is cited as proof that “lowering sodium lowers calcium and prevents stones.” The data are subtler: urinary calcium fell by essentially the same amount (approximately 170 mg/day) in **both** arms. What separated them was oxalate — it rose in the low-calcium group and fell in the normal-calcium group. The benefit was substantially an oxalate effect mediated by maintained calcium intake, not a clean demonstration that sodium restriction prevents stones by lowering calciuria^([3](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-3)).

### 8.2 The trialists themselves flagged the bias

Two foundational positive trials contain warnings from their own authors. In the allopurinol trial, the placebo group had **63% fewer** calculi than its own historical baseline; the authors wrote this “underscores the positive treatment bias that regularly occurs in trials of prophylaxis against renal calculi when historical controls are used”^([7](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-7)). The chlorthalidone trial made the point more forcefully — every arm, including placebo, showed large reductions from predicted rates, “demonstrat\[ing\] the need for proper experimental design”^([6](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-6)). These are confessions embedded in the primary literature.

## 9. Why the Evidence Is Weaker Than It Looks

- **The stone clinic effect / regression to the mean.** Patients enroll after a cluster of stones — a statistical peak that regresses regardless of treatment. Hosking 1983 showed 58% of idiopathic calcium formers became metabolically inactive on conservative fluid and diet advice alone, and recommended withholding drugs until this effect is assessed^([13](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-13)). Trials using pre-treatment or historical control rates attribute this spontaneous regression to the drug.
- **Small, old, single-center trials.** The positive citrate, allopurinol, and thiazide trials enrolled dozens to low-hundreds at single centers decades ago, with composite (often asymptomatic, imaging-detected) endpoints that inflate effect sizes.
- **The modern trials are negative.** With large, double-blind, placebo-controlled designs, both flagship results were null — NOSTONE (a companion analysis also found no bone-density benefit)^([4](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-4),[5](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-5)) and PUSH^([1](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-1)). Citrate and allopurinol have never had a NOSTONE-scale modern replication.
- **Surrogate-driven reasoning.** Much of practice optimizes 24-hour numbers and supersaturation rather than demonstrated event reduction. NOSTONE is the cautionary tale: thiazide reliably lowers urinary calcium, yet that produced no reduction in recurrence^([4](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-4)). The RNAi oxalate trials, however rigorous, are also surrogate-endpoint studies^([14](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-14)).

Clinical Pearl

This does not mean nothing works. Fluid and citrate retain plausibility and some positive (if dated) randomized signal, and stone chemistry is real. It means confidence intervals are wide, the best modern trials humble the older ones, and honest practice treats these as reasonable, low-harm, mechanism-based interventions — not proven, precisely-quantified therapies.

## 10. PUSH and NOSTONE: The Two First-Line Beliefs That Failed Their Best Trials

Two interventions are recited as the foundation of stone prevention — **thiazides** for hypercalciuria and **water** for everyone. Each was finally tested in a large, modern, well-run randomized trial, and each came back null. Because these two beliefs anchor most counseling, they deserve a closer look than a single table row.

### 10.1 NOSTONE — thiazide, at three doses, did nothing

NOSTONE randomized 416 recurrent calcium stone formers to hydrochlorothiazide 12.5, 25, or 50 mg/day or placebo over a median 2.9 years^([4](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-4)). The primary composite (symptomatic or radiographic recurrence) was **essentially identical across all four arms — no benefit at any dose and no dose-response** — even though thiazide did what it is supposed to do and lowered urinary calcium. This is the cleanest cautionary tale in the field: **a therapy can reliably move the surrogate (calciuria) and still not move the outcome.** A companion analysis found no bone-density benefit either^([5](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-5)). The older positive thiazide trials that built the guideline recommendation were small, single-center, and decades old.

### 10.2 PUSH — an intensive program raised urine volume but not stone-free rates

PUSH (Prevention of Urinary Stones with Hydration) was an NIDDK / Urinary Stone Disease Research Network RCT of a *behavioral adherence intervention*, not of water itself^([1](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-1)). Adults and adolescents (≥12 years) with a symptomatic-stone history and low urine volume were randomized 1:1; **both** arms received usual care plus a smart water bottle and smartphone app, while the intervention arm added a fluid prescription, **financial incentives** (approximately \$1.50/day for hitting the fluid goal), health coaching, and structured problem-solving. The design and rationale were pre-published^([15](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-15)).

**Primary result — negative.** At a median 2-year follow-up, symptomatic stone events were **154 (19%) with the intervention vs 165 (20%) with control (HR 0.96, 95% CI 0.77–1.20)**. There was no difference in stone growth of at least 2 mm, in new radiographic stones, or in the composite of recurrence/new-stone/growth. On safety, no hyponatremia required hospitalization; asymptomatic hyponatremia occurred in 12 (1%) intervention vs 2 (\<1%) control participants, and urinary storage symptoms (frequency, urgency, nocturia) were transiently greater in the intervention arm at months 6 and 12.

The achieved-volume problem — they never delivered the dose

The revealing number is not the arm assignment but the urine volume actually achieved. Across the adult cohort, median 24-hour volume rose from approximately 1.3 to approximately 1.8 L/day — a median change of roughly **+438 mL/day** — and **fell well short of the 2.5 L/day guideline target the program was designed to hit.** Despite a fluid prescription, daily financial incentives, coaching, and smart bottles, most participants stayed in the 1.5–2 L range. PUSH tried to estimate a dose-response from a trial that never delivered the dose. Critically, pre-specified analyses showed no heterogeneity by adherence — **even the subset randomized to intervention who did reach ≥2.5 L/day showed no signal.**

### 10.3 Achieved urine volume and exploratory evidence

A secondary PUSH analysis presented at AUA 2026 examined achieved urine volume and osmolality. This is conference-level evidence, rather than a separate randomized comparison of achieved hydration targets. It should be considered exploratory.

Achieved urine volume is measured after randomization and may be associated with adherence, diet, underlying disease, and other patient characteristics. Associations based on achieved values therefore do not establish the causal benefit of a particular volume or osmolality target. Interpret the randomized intervention result separately from these secondary associations.

Clinical interpretation

The primary PUSH trial did not demonstrate that its behavioral hydration program reduced symptomatic stone recurrence. This does not by itself establish that all hydration strategies are ineffective. Continue individualized fluid and dietary counseling, assess the patient's ability to follow the plan, and avoid promising a precise risk reduction.

The primary randomized report is [Desai et al., Lancet 2026](https://pubmed.ncbi.nlm.nih.gov/41864748/). The AUA 2026 secondary analysis is listed separately as a conference abstract.

## 11. Practical Synthesis

- Get stone composition and two 24-hour collections in recurrent/high-risk formers; read supersaturations as the dashboard and verify adequacy by creatinine.
- Lead with fluid and diet (normal calcium, lower sodium and animal protein, oxalate moderation, calcium *with* meals if enteric hyperoxaluria) — low harm, plausible, the only interventions even arguably supported in single-stone formers — and give them a real trial before drugs (the stone clinic effect)^([13](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-13)).
- Reserve pharmacotherapy for genuinely recurrent/active disease: potassium citrate for hypocitraturia and uric-acid stones, allopurinol only when hyperuricosuria is documented, thiazide for hypercalciuria — while being candid that NOSTONE weakens the thiazide case^([4](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-4)).
- Do not reach for magnesium (no monotherapy RCT support)^([6](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-6)) and be cautious with citrate in calcium-**phosphate** disease (it can raise CaP supersaturation; monitor pH/SS).
- Treat CaP and struvite as their own problems: evaluate CaP for hyperparathyroidism/RTA; manage struvite surgically with organism eradication^([12](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-12)). Refer suspected primary hyperoxaluria for genotyping and RNAi therapy^([14](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html#ref-14)).
- Track response with follow-up 24-hour urine / supersaturation and imaging, recognizing you are monitoring a surrogate — and that no calcium-stone therapy has modern randomized proof of halting growth of an existing stone.

Clinical Pearl

The honest framing for patients and trainees: “We have a sound physical-chemistry model and decades of mostly small trials suggesting these measures help; the largest modern trials have been disappointing; so we use low-harm, mechanism-based therapy and measure your response, rather than promising a precise risk reduction.”

## References

1.  Desai AC, Maalouf NM, Harper JD, et al. Prevention of urinary stones with hydration (PUSH): a randomised clinical trial of an adherence intervention. *Lancet.* 2026;407(10534):1171–1181. [PMID: 41864748](https://pubmed.ncbi.nlm.nih.gov/41864748/)
2.  Borghi L, Meschi T, Amato F, et al. Urinary volume, water and recurrences in idiopathic calcium nephrolithiasis: a 5-year randomized prospective study. *J Urol* 1996. [PubMed 8583588](https://pubmed.ncbi.nlm.nih.gov/8583588/).
3.  Borghi L, Schianchi T, Meschi T, et al. Comparison of two diets for the prevention of recurrent stones in idiopathic hypercalciuria. *N Engl J Med* 2002. [PubMed 11784873](https://pubmed.ncbi.nlm.nih.gov/11784873/).
4.  Dhayat NA, Bonny O, Roth B, et al. Hydrochlorothiazide and Prevention of Kidney-Stone Recurrence. *N Engl J Med* 2023. [PubMed 36856614](https://pubmed.ncbi.nlm.nih.gov/36856614/).
5.  Christe A, Primetis E, Cereghetti GM, et al. Hydrochlorothiazide and Bone Mineral Density in Patients with Kidney Stones: Post Hoc Analysis of the NOSTONE Trial. *Clin J Am Soc Nephrol* 2025. [PubMed 40063114](https://pubmed.ncbi.nlm.nih.gov/40063114/).
6.  Ettinger B, Citron JT, Livermore B, et al. Chlorthalidone reduces calcium oxalate calculous recurrence but magnesium hydroxide does not. *J Urol* 1988. [PubMed 3280829](https://pubmed.ncbi.nlm.nih.gov/3280829/).
7.  Ettinger B, Tang A, Citron JT, et al. Randomized trial of allopurinol in the prevention of calcium oxalate calculi. *N Engl J Med* 1986. [PubMed 3534570](https://pubmed.ncbi.nlm.nih.gov/3534570/).
8.  Barcelo P, Wuhl O, Servitge E, et al. Randomized double-blind study of potassium citrate in idiopathic hypocitraturic calcium nephrolithiasis. *J Urol* 1993. [PubMed 8230497](https://pubmed.ncbi.nlm.nih.gov/8230497/).
9.  Ettinger B, Pak CY, Citron JT, et al. Potassium-magnesium citrate is an effective prophylaxis against recurrent calcium oxalate nephrolithiasis. *J Urol* 1997. [PubMed 9366314](https://pubmed.ncbi.nlm.nih.gov/9366314/).
10. Fink HA, Wilt TJ, Eidman KE, et al. Medical management to prevent recurrent nephrolithiasis in adults: a systematic review for an American College of Physicians Clinical Guideline. *Ann Intern Med* 2013. [PubMed 23546565](https://pubmed.ncbi.nlm.nih.gov/23546565/).
11. Pearle MS, Goldfarb DS, Assimos DG, et al. Medical management of kidney stones: AUA guideline. *J Urol* 2014. [PubMed 24857648](https://pubmed.ncbi.nlm.nih.gov/24857648/).
12. Griffith DP, Gleeson MJ, Lee H, et al. Randomized, double-blind trial of Lithostat (acetohydroxamic acid) in the palliative treatment of infection-induced urinary calculi. *Eur Urol* 1991. [PubMed 1726639](https://pubmed.ncbi.nlm.nih.gov/1726639/).
13. Hosking DH, Erickson SB, Van den Berg CJ, et al. The stone clinic effect in patients with idiopathic calcium urolithiasis. *J Urol* 1983. [PubMed 6644890](https://pubmed.ncbi.nlm.nih.gov/6644890/).
14. Garrelfs SF, Frishberg Y, Hulton SA, et al. Lumasiran, an RNAi Therapeutic for Primary Hyperoxaluria Type 1. *N Engl J Med* 2021. [PubMed 33789010](https://pubmed.ncbi.nlm.nih.gov/33789010/).
15. Scales CD, Desai AC, Harper JD, et al; Urinary Stone Disease Research Network. Prevention of Urinary Stones With Hydration (PUSH): design and rationale of a clinical trial. *Am J Kidney Dis.* 2021;77(6):898-906.e1. [PMID: 33212205](https://pubmed.ncbi.nlm.nih.gov/33212205/)
16. Coronary Drug Project Research Group. Influence of adherence to treatment and response of cholesterol on mortality in the Coronary Drug Project. *N Engl J Med.* 1980;303(18):1038-1041. [PMID: 6999345](https://pubmed.ncbi.nlm.nih.gov/6999345/)
17. Harper J, Lieske J, Desai A, et al. Urine volume and urine osmolality and risk of recurrent stone events: results from the Prevention of Urinary Stones With Hydration (PUSH) trial. *J Urol*. 2026;215(5S2):e6. [AUA conference abstract (not a full peer-reviewed PubMed article)](https://doi.org/10.1097/01.JU.0001192572.07890.f8.13).

*Effect sizes are quoted as reported in the primary articles and the AHRQ/ACP systematic review. Statements of “no RCT” reflect the absence of randomized event-outcome trials for that specific indication as of June 2026.*


---

[Website version](https://urinenephrology.org/mastery/stone-disease/metabolic-stone-workup.html) · Markdown synchronized October 3, 2026.