Kidney Stones and the Risk of Coronary Heart Disease

Abstract

This article showed that women, but not men, with a history of kidney stones had a higher rate of coronary heart disease (CHD; defined by fatal or nonfatal myocardial infarction or coronary revascularization) than comparable participants without a history of stones. Starting with participants with and without a history of kidney stones and without a history of CHD, alargenumberofpeople—morethan45,000menfrom the Health Professionals Follow-up Study (HPFS) and more than 196,000 women from the Nurses’ Health Study (NHS) I and II—were followed prospectively for a median of 9 years. With over 1.3 million personyears of follow-up, individuals with a history of stone disease in the older cohort of women (NHS I, mean age,59years)hadamultivariablehazardratio(HR)for development of total CHD of 1.18 (95% confidence interval [CI], 1.08-1.28) compared with women without stones; the younger cohort of women with stones(NHSII, mean age,37.4years), with 1.8million person-years of follow-up, had a multivariable HR of 1.48 (95% CI, 1.23-1.78). Men with kidney stones drawn from the HPFS, where the mean age was 55.8 years, were no more likely to develop CHD than individuals who did not form stones, as seen in 0.8 million person-years of follow-up, with multivariable HR of 1.06 (95% CI, 0.99-1.13). Much of the knowledge gained in the last 20 years regarding the epidemiology of stones arises from these 3 cohorts, which have been followed by Curhan, Taylor, and co-investigators. There are many reasons these cohorts remain a reliable and continuous source of useful information about kidney stones. The cohorts are large, offering significant power to detect even weak associations of stones with other conditions. In other studies, participants with no prior history of stones could be followed to determine variables associated with the development of stones. For example, one seminal study of the HPFS over 20 years ago demonstrated that greater intake of dietary calcium was associated with fewer stones, 2 while a second study, published nearly a decade ago, showed that the presence of obesity and weight gain increase the likelihood of stones. 3 These reports are all prospective observational studies, with both the strengths and limitations of that study design. Importantly, both the data regarding diagnoses, such as myocardial infarction, and variables, such as dietary intake, have been thoroughly validated. For instance, self-reported kidney stones were shown to be reported accurately in 97% of cases. 4 In the current study, myocardial infarction or coronary heart disease were carefully documented by state death certificates, National Death Index data, autopsy records, next of kin, and medical records, while self-reported revascularization was shown to be “virtually 100% specific,” likely in part reflecting that all participants are health care workers. 5 While the strengths are abundant, the weaknesses of these observational studies are inherent to the nature of the follow-up. These studies require multivariable adjustments that may fail to account for unanticipated or unknown confounders. In this case, perhaps some unknown variable or a known variable that was insufficiently accounted for is common to stone formers but not to non–stone formers and is responsible for the association between stone formation and increased development of coronary disease. The authors repeatedly caution that these cohorts are not representative of all populations in the United States or elsewhere, and that the participants, being motivated health professionals, may be composed of racial, socioeconomic, or geographic groupings that influence the findings. Finally, these observational studies reveal associations and not causal relationships; they must be viewed as hypothesis-generating investigations that could lead to interventional studies. Unfortunately, appropriate investigations have not always followed, though an important