Kidney Stones: What the Prevention Evidence Actually Supports

If you have passed one kidney stone, the practical question is how to avoid the next one, and the honest answer is that the confident-sounding advice rests on thinner evidence than most people expect. One habit, drinking enough water to keep urine dilute, has broad support across guidelines but only a weak grade behind it. The medicines that get prescribed, thiazide diuretics, potassium citrate, and allopurinol, do have randomized backing, but mostly in narrow groups of patients whose urine chemistry points to a specific problem. The test that sorts a targeted plan from an educated guess is a 24-hour urine collection.

Key points#

• Fluid intake enough to produce at least 2 to 2.5 liters of urine a day is the one prevention step both major guidelines endorse, yet the American College of Physicians (ACP) still graded it a weak recommendation on low-quality evidence.
• The single trial doing most of the work behind that advice, from 1996, cut five-year recurrence roughly in half in first-time calcium stone formers.
• A 24-hour urine collection tells you whether calcium, citrate, oxalate, or uric acid is the driver, and that is what makes drug choice rational.
• Thiazides, citrate, and allopurinol each have randomized support, but chiefly within the metabolic subgroup each targets.
• A large 2023 trial (NOSTONE) found hydrochlorothiazide no better than placebo when given to recurrent stone formers broadly, which sharpens the case for testing before treating.

Start with the fork in the road: the urine collection#

Most disagreement about stone prevention dissolves once you separate two situations. For a first, uncomplicated calcium stone, guidelines lean pragmatic: the ACP does not require metabolic testing before advising more fluid, since hydration is cheap and low-risk. For recurrent stones, or a first stone in someone at high risk, the AUA guideline recommends a fuller metabolic evaluation built on one or two 24-hour urine collections. The sample is analyzed at minimum for volume, pH, and calcium. It is analyzed for oxalate, uric acid, and citrate. It is analyzed for sodium, potassium, and creatinine.

Think of that collection as a chemistry readout of the fluid your stones actually form in. High urine calcium, low citrate, and a uric-acid-driven pattern each call for a different medicine. Skip the collection and drug selection drifts toward habit rather than the patient in front of you. That is the thread running through everything below: the drugs are only as good as the target they are matched to.

What "drink more water" is really built on#

The recurring recommendation to stay well hydrated traces back largely to a single study. In 1996, Borghi and colleagues randomized 199 adults after a first idiopathic calcium stone either to high water intake, aimed at producing more than 2 liters of urine daily, or to no specific fluid instruction, and followed them for five years. Recurrence was about 12 percent with high intake versus 27 percent in the comparison group, and the stones that did form took longer to appear.

That is a real and useful result, but it is one trial in mostly first-time calcium stone formers. The ACP guideline reflects those limits honestly: its lead recommendation to increase fluid to reach at least 2 liters of urine daily is graded weak, on low-quality evidence. The advice is sensible and carries little downside, which is why it is offered so widely. It is worth noticing, though, that a broadly recommended step and a strongly proven one are not the same thing.

The drugs, and the subgroup each one fits#

Thiazides, and the trial that complicated them#

Older randomized trials of thiazide-type diuretics showed reduced recurrence in recurrent stone formers, often those with high urine calcium, and the ACP graded drug monotherapy a weak recommendation on moderate-quality evidence. Then came the NOSTONE trial, published in the New England Journal of Medicine in 2023, which randomized 416 recurrent calcium stone formers to hydrochlorothiazide at 12.5, 25, or 50 mg, or to placebo, for about three years. Recurrence did not differ meaningfully across the groups, and there was no clear dose response.

The likely reason is in the enrollment: NOSTONE took recurrent formers broadly rather than only those with high urine calcium, the very group thiazides were expected to help. The trial does not erase the earlier evidence so much as it warns against assuming a thiazide prevents stones regardless of whether the metabolic target is present.

Citrate, for low urinary citrate#

Potassium citrate has randomized support in patients with low urinary citrate, where trials showed fewer recurrences, and the AUA offers it to that group. It also raises urinary pH, which matters for uric acid and cystine stones. Its case is strongest when a collection documents low citrate and weakest as a blanket measure handed to everyone.

Allopurinol, for one specific phenotype#

Allopurinol has the narrowest evidence of the three. A randomized trial found it reduced calcium oxalate recurrence specifically in people with high urinary uric acid and normal urine calcium, which is exactly the phenotype the AUA targets. Outside that combination, its benefit is not established.

Diet: broadly safe, and one counterintuitive finding#

Dietary advice is where broad, low-risk habits do apply widely: drink more, moderate sodium and animal protein, and, importantly, keep dietary calcium normal rather than low. That last point trips people up, because it seems logical that calcium stones would call for less calcium. A separate randomized trial in men with recurrent stones and high urine calcium pointed the other way, finding that a diet with normal calcium but reduced sodium and animal protein produced fewer recurrences than the older low-calcium approach. Cutting dietary calcium can actually free up more oxalate to be absorbed, which is counterproductive.

The honest summary#

Two rules of thumb hold up. First, low-risk habits, hydration and sensible diet, apply to nearly everyone and align with both guidelines. Second, drug therapy is subgroup-specific: its randomized support lives inside defined urine phenotypes, which is why the AUA anchors it to urine chemistry, and why the ACP found that combining agents did not beat monotherapy, so more pills is not more prevention. The durable picture is one modest but real habit and a small set of drugs that work in defined niches, with the boundaries of those niches drawn by a urine collection rather than by intuition.