ation of quality of salt studies. Sample size, number of events, and participation and drop-out rates in our study are similar to those in most previous reports. As highlighted by Cook, the strongest evidence comes from randomized clinical trials rather than observational studies. In the Trials of Hypertension Prevention, participants collected up to 7 urine samples over 3 years, but the risk of a cardiovascular event across quartiles was unrelated to sodium excretion over time. A Cochrane Review demonstrated that the relative risk of all-cause mortality associated with salt reduction was 0.90 (95% CI, 0.58-1.40) in normotensive participants and 0.96 (95% CI, 0.83-1.11) in hypertensive patients. Thus, currently available randomized clinical trials do not contradict our longer-term follow-up data.