HDCN Article Review/Hyperlink

Krieger JN, Kronmal RA, Coxon V, Wortley P, THompson L, Sherrard DJ

Dietary and behavioral risk factors for urolithiasis: potential implications for prevention

Am J Kidney Dis (Aug) 28:195-201 1996

Much data exist suggesting that diet is an important risk factor for stone formation (1). More affluent members of industrialized societies consume diets containing greater amounts of animal protein and purines are at increased risk, as are those who ingest greater amounts of sodium and oxalates (2). Fluid intake appears to be an important determinant of stone risk, perhaps explaining the increased incidence of stone formation in "stone belts" like the southern United States and the Middle East (1). Controlled trials in which these variables are manipulated prospectively though have very rarely been done. We rely therefore on retrospective data to study risk factors for nephrolithiasis.

The authors of the current study have proceeded in this tradition by performing a prospective case-control study of risk factors for stone formation in a large managed care organization in Seattle. They legitimately claim that their data have special merit since they arise from a "primary care" setting and not the usual referred population seen in stone centers. The cases were all men between 25 and 55 experiencing their first episode of stone formation. The self-contained nature of managed care organizations makes it credible that case-finding was complete. Patients with previous episodes of stones were excluded. Controls, men matched for age, without history of stone formation, were difficult to obtain, and the authors are forthcoming about this problem. In the end they had 240 cases and 392 controls (only 22% of those eligible). They then performed telephone interviews collecting demographic variables, and dietary and medication variables possibly related to stone formation.

Reduced risk for stone disease was associated with postsecondary education, low-fat or weight loss diet, and antacid use. Beer drinking was protective, associated with a relative risk for stones of 0.44, with greater amounts of beer ingested not giving further protection. Other drinks (coffee, tea or soda) were not associated with altered risk. Logistic regression confirmed low-fat or weight reduction diet, beer drinking and antacid use as being protective. African-American ethnicity, postsecondary education, and a warm work place were also protective. A positive family history of stones and diuretic use both conferred a doubling of risk for stones. These associations all appear to be statistically significant, with impressive p values, and reasonable confidence intervals.

Comment: The first issue in case-control studies is the selection of the controls. There is always the chance that the controls are a self-selected group that somehow differ from the general population. In this case, other than the relatively low percentage of controls successfully recruited, one has no clear basis for finding that the controls are not appropriately selected. The larger problem here is the biologic plausibility: why should these variables be important and influence stone formation in the direction they do? The authors can offer hypotheses for their findings, but do so in a rather post-hoc fashion. This is particularly suggested by the authors' statement that the data were collected as part of an investigation of a correlation between vasectomy and stone formation. This detracts from the prospective nature of the trial. The data therefore may represent true determinations of risk factors, and may serve as the basis for further research, but there is also always the danger of asking too many questions and producing Type I errors. Without prospective delineation of the details of the variables studied and found to vary from cases to controls, our understanding of the relevant pathophysiology is not necessarily advanced.

In the case of beer drinking, the authors' finding would not ordinarily be expected since beer is often cited as a beverage with relatively high oxalate content. Its protective effect however has been demonstrated in a previous study (showing relative risk of 0.36- 0.58)(3), adding credibility to the present finding. Drinking more beer added no further protection, and drinking other liquids was not protective, so no ready explanation is at hand. Perhaps oral oxalate reduces calciuria. Though this may seem counter-intuitive, oral ingestion of oxalate in hypercalciuric rats reduced urinary calcium oxalate supersaturation (4).

It is reassuring that a family history conferred a higher risk, but disturbing that diuretic use increased risk as well. These presumably relatively healthy, relatively young men, should not be taking much furosemide, a drug that increases calciuria. Instead, one would expect them to be taking thiazides for hypertension, a class of drugs definitively shown to reduce stone formation (5). As the authors point out, thiazides may cause hypocitraturia (via hypokalemia) and triamterene has low solubility and can cause stones itself. But the authors did not ascertain the nature of the diuretic use, and the number of patients taking them was obviously small. Similarly the authors must have prospectively believed that a warm work place would increase risk; yet it reduced risk. This is unexpected and varies from previous data.

The authors attribute the improved risk of low-fat or weight loss diet to reduction in protein intake, but that this is what occurred is not documented. (The trendy weight loss diet in New York in the '90s is HIGH protein). The benefits of antacids may be due to increased urinary pH, oxalate binding, or magnesium supplementation (though most trials of magnesium supplements have not demonstrated efficacy) (5). But the type of antacid and the indication for them was not determined.

In summary, the authors have demonstrated one advantage of studying the self- contained managed care environment, by determining several risk factors in a primary care population. The questions raised, but not answered, about pathophysiologic mechanisms are important and should lead to fruitful research. Two studies have now shown a protective effect of beer: it should now be struck from the list of forbidden liquids where it occasionally appears. That Seattle is at the forefront of the American trend favoring microbrews makes this group's finding particularly apropos. Zalman Agus said at the ASN this past year that a beer and acetazolamide at bedtime was an excellent therapy for uric acid stones. As a homebrewer, and stone-former, I am tremendously gladdened by the transformation of beer into a therapeutic libation. Cheers! David S. Goldfarb, M.D. Co-Director, Kidney Stone Prevention and Treatment Program, New York University School of Medicine

1. Asplin, J., Chandhoke, P.S. The stone-forming patient, in Kidney Stones. Medical and Surgical Management, Coe, F.L., Favus, M.J., Pak, C.Y.C, Parks, J.H., Preminger, G.M., eds., Lippincott-Raven, New York, 1996.

2. Goldfarb, S. Dietary factors in the pathogenesis and prophylaxis of calcium nephrolithiasis. Kidney Int 34:544-555 (88).

3. Shuster, J., Finlayson, B., Scheaffer, R.L., Sierakowski, R., Zoltek, J., Dzegede, S. Primary liquid intake and urinary stone disease. J Chron Dis 39: 907-914 (85).

4. Bushinsky, D., ... Coe, F.L. J Am Soc Nephrol (95).

5. Ettinger, B., Citron, J.T., Livermore, B., Dolman, L.I. Chlorthalidone reduces calcium oxalate calculous recurrence but magnesium hydroxide does not. J Urol 139:679-684 (88).