Sodium and cardiovascular disease.

To the Editor: In 2013, I served on an Institute of Medicine (IOM) panel whose members concluded that data published through 2012 relating cardiovascular outcomes to a sodium intake below 2300 mg per day were inconsistent and insufficient.1 The article by O’Donnell and colleagues (Aug. 14 issue),2 which presents data from the Prospective Urban Rural Epidemiology (PURE) study, only adds to the confusion. A single morning urine sample is an inaccurate measure of usual sodium intake, ignoring day-to-day variability in sodium intake, diurnal variation in sodium excretion, and the effects of medication.3 The Kawasaki formula also overestimates sodium exposure in the U.S. population.4 In contrast, in the Trials of Hypertension Prevention (TOHP),5 an average of three to seven 24-hour urine collections were performed over 1 to 4 years to derive a very accurate estimate of a person’s typical sodium intake. After the IOM report was released, the TOHP data showed a direct, progressive relationship between lower sodium intake (down to intake below 1500 mg per day) and lower cardiovascular risk and there was no indication of a J-shaped curve.5 The fact that the PURE study is the largest to date should not influence its interpretation. A large study size does not eliminate bias resulting from selection, reverse causation, or confounding but could lead to spurious results, with a very small confidence interval around a very biased estimate. Nancy R. Cook, Sc.D.