Kidney stones are a global healthcare problem and both incidence and prevalence of renal stone disease are increasing, irrespective of age, sex and race.1 In the US, the current lifetime risk of a symptomatic stone event is ~18.8% in men and ~9.4% in women.2 Without a specific treatment, 5- and 20-year recurrence rates are ~40 and ~75%, respectively.3 Both direct (hospitalizations, surgery) and indirect (lost work time) cost associated with kidney stones are enormous and the morbidity related to recurrent kidney stone disease is high. Hence, medical prophylaxis seems to be an attractive approach. Indeed, apart from its benefits to patients in terms of reduced morbidity and risk from procedures, medical prevention of nephrolithiasis is cost effective.4

Most kidney stones are composed of calcium oxalate, calcium phosphate or a mixture of both. Although various inherited and systemic diseases are associated with kidney stones, most stones are idiopathic. Increased excretion of calcium in the urine, hypercalciuria, is the most frequent metabolic abnormality encountered in patients with recurrent idiopathic calcareous nephrolithiasis. Thiazide diuretics have been the cornerstone of pharmacologic stone metaphylaxis since 50 years. The effect of thiazides to reduce the risk of stone recurrence has been attributed to their ability to decrease urinary calcium excretion. However, other factors, such as reduction of urinary pH and urinary oxalate excretion, probably contribute to this effect. The efficacy of thiazides on recurrence prevention of calcareous nephrolithiasis was tested in several small RCTs in the 1980s and 1990s.5 With the exception of two trials, thiazides significantly reduced stone recurrence in recurrent stone formers, in average by ~50%. A recent meta-analaysis revealed major methodological deficiencies in these trials, including lack of double-blinding and intention-to-treat analysis, unclear allocation concealment and lack of adverse event and drop out reporting.6 In addition to methodological concerns, high doses of thiazides were employed in these trials, in the case of the best studied thiazide, hydrochlorothiazide, 25 or 50 mg twice daily. At such high doses, side effects occur in 30–40% of patients and include electrolyte disturbances, gout, muscle cramps, impaired glucose tolerance, gastrointestinal symptoms and erectile dysfunction.7 Nowadays, thiazides are widely used in the treatment of both recurrent nephrolithiasis and arterial hypertension, but at significantly lower doses.5 In the case of recurrent nephrolithiasis, however, this practice is not supported by randomized evidence and consequently, we do not know whether the currently employed low-dose thiazide regimens are effective in reducing the risk for stone recurrence. In addition to thiazides, calcium channel blockers and ACEi/ARBs are widely used antihypertensive medications. Both calcium channel blockers and ACEi/ARB increase urinary calcium excretion and thereby may cause an excess of kidney stone events in hypertensive patients. If these medications indeed increase the risk of stone formation is currently unclear.

In this issue, Alexander et al.8 report the results of a population-based observational cohort study, where they studied the impact of different antihypertensives on kidney stone risk. To this end, the authors employed linked administrative health databases of the State Ontario, Canada. Hypertensive patients aged >65 years with a new prescription of antihypertensive medication were included in the study. The primary outcome of the study was occurrence of a kidney stone event, median follow-up time was 368 days. Results were adjusted for various co-variates, including age, sex and the comorbidities diabetes, arterial hypertension, kidney stone disease and chronic inflammatory bowel disease. Compared to beta blockers, thiazide medication was associated with a significantly lower risk of a stone event, ACEis/ARBs were associated with a borderline decreased risk, while calcium channel blockers were associated with a comparable risk.

The study has the usual shortcomings of a retrospective study design, and despite adjustment for many co-variates, residual confounding remains a concern. Further limitations of the study are the very short follow-up time and the fact that only hypertensive individuals aged >65 years were included. Also, incidence of side effects (electrolyte disturbances, worsening of renal function, glucose intolerance) and need for hospitalization were not assessed. A surprising finding is the high incidence of symptomatic stone events observed in the cohort (0.83% per year). The reason for this is not entirely clear but maybe at least partially due to the fact that older patients with arterial hypertension were studied. Both old age and arterial hypertension are associated with a significantly increased prevalence of kidney stone disease.2, 9 The significant thiazide effect observed during the short follow-up time and the fact that renal calculi typically take months to years to form suggests that thiazides not only reduce stone formation but also stone passage. Unfortunately, type and dose of thiazides prescribed were not analyzed in the study. Since thiazides were primarily prescribed for blood pressure control, it seems reasonable to assume that the great majority of patients were on a low-dose thiazide regimen mainly in form of hydrochlorothiazide, the most frequently prescribed thiazide.

Despite the limitations mentioned, the impressive number of patients (542 581) included in the analysis supports the robustness of the findings and make this a valuable study for the clinical community. Reassuringly, calcium channel blockers and ACEi/ARBs were not associated with an increased risk of symptomatic stone events compared to beta blockers, at least with short term use. Moreover, thiazides seem to be effective in preventing symptomatic kidney stone events in elderly patients with arterial hypertension, extending previous observations in RCTs with recurrent stone formers. Apart from being effective blood pressure-lowering drugs and their well-known beneficial effects on bone health,10 reduction of kidney stone events seems to be yet another reason to prescribe thiazides in elderly hypertensive patients, even in the absence of a history of kidney stone disease.

Clearly, studies with longer follow-up time and inclusion of younger patients are now needed to substantiate the findings and determine the generalizability of the results. In addition, we need more information with respect to type and dose of thiazides employed. Thiazides are characterized by a flat dose response with respect to blood pressure reduction, but their side effects are dose dependent.11, 12, 13 Unfortunately, pharmacodynamics of thiazides for stone prevention have not been studied until now. This leaves the physician who treats a recurrent stone former with arterial hypertension (not a rare species) in a classical dilemma: prescribe a high thiazide dose and accept an increased risk of potentially serious side effects or prescribe a low thiazide dose, which is not supported by clinical evidence. The study by Alexander et al.8 is a clear wake-up call: it is time that we address this important clinical problem with prospective trials.