Evaluation of a brief sodium screener in two samples

Abstract

The Sodium Screener© (SS©), as developed by NutritionQuest (Berkeley, CA, USA), was designed to reduce the burden of repeated dietary or urinary sodium measurements, but the accuracy of daily sodium intake estimates has not been reported. Associations were examined between sodium intakes derived from the SS© scores and repeated 24-h recalls (24DR) in two studies with different administration modes. In one study, 102 registered dietitians (RD) completed three Automated Self-Administered 24DRs (ASA24©), version 2014, followed by the SS©; both were self-administered and web-based. In the second sample, (the Study of Household Purchasing Patterns, Eating, and Recreation or SHoPPER), trained dietitians conducted 24DR interviews with 69 community-dwelling adults in their homes; all the community adults then completed a paper-based SS© at the final visit. In the RD study, SS©-predicted sodium intakes were 2604 ± 990 (mean ± Standard deviation (SD)), and ASA24© sodium intakes were 3193 ± 907 mg/day. In the SHoPPER sample, corresponding values were 3338 ± 1310 mg/day and 2939 ± 1231 mg/day, respectively. SS©-predicted and recall sodium estimates were correlated in the RD study (r = 0.381, p = 0.0001) and in the SHoPPER (r = 0.430, p = 0.0002). Agreement between the SS© and 24-h recalls was poor when classifying individuals as meeting the dietary sodium guidelines of 2300 mg/day or not (RD study: kappa = 0.080, p = 0.32; SHoPPER: kappa = 0.207, p = 0.08). Based on repeated 24DR either in person or self-reported online as the criterion for estimating daily sodium intakes, the SS© may require additional modifications.