Use of a Single Baseline Versus Multiyear 24-Hour Urine Collection for Estimation of Long-Term Sodium Intake and Associated Cardiovascular and Renal Risk

Abstract

Background: A decrease in sodium intake has been shown to lower blood pressure, but data from cohort studies on the association with cardiovascular and renal outcomes are inconsistent. In these studies, sodium intake was often estimated with a single baseline measurement, which may be inaccurate considering day-to-day changes in sodium intake and sodium excretion. We compared the effects of single versus repetitive follow-up 24-hour urine samples on the relation between sodium intake and long-term cardiorenal outcomes. Methods: We selected adult subjects with an estimated glomerular filtration rate >60 mL/min/1.73m 2 , an outpatient 24-hour urine sample between 1998 and 1999, and at least 1 collection during a 17-year follow-up. Sodium intake was estimated with a single baseline collection and the average of samples collected during a 1-, 5-, and 15-year follow-up. We used Cox regression analysis and the landmark approach to investigate the relation between sodium intake and cardiovascular (cardiovascular events or mortality) and renal (end-stage renal disease: dialysis, transplantation, and/or >60% estimated glomerular filtration rate decline, or mortality) outcomes. Results: We included 574 subjects with 9776 twenty-four–hour urine samples. Average age was 47 years, and 46% were male. Median follow-up was 16.2 years. Average 24-hour sodium excretion, ranging from 3.8 to 3.9 g (165–170 mmol), was equal among all methods ( P =0.88). However, relative to a single baseline measurement, 50% of the subjects had a >0.8-g (>34-mmol) difference in sodium intake with long-term estimations. As a result, 45%, 49%, and 50% of all subjects switched between tertiles of sodium intake when the 1-, 5-, or 15-year average was used, respectively. Consequently, hazard ratios for cardiorenal outcome changed up to 85% with the use of sodium intake estimations from short-term (1-year) and long-term (5-year) follow-up instead of baseline estimations. Conclusions: Relative to a single baseline 24-hour sodium measurement, the use of subsequent 24-hour urine samples resulted in different estimations of an individual’s sodium intake, whereas population averages remained similar. This finding had significant consequences for the association between sodium intake and long-term cardiovascular and renal outcomes.