Comparison of analytical methods to determine sodium content of low-sodium foods

U.S. nutrition labeling regulations require the declaration of sodium content on food products. Accurate and reproducible determination of Na in foods with low Na content (< 140 mg/serving) is challenging because of laboratory contamination. Within-laboratory performance of inductively coupled plasma/MS (ICP/MS), flame atomic absorption spectrophotometry (FAAS), ion-selective electrode (ISE), and potentiometric titration of chloride ion were evaluated in 17 low-sodium foods. For 13 types of food, statistically significant differences (P < 0.05) exist between the within-day andlor interday means obtained by ICP/MS, FAAS, and ISE. Median within-day and interday precent RSD values were 2.7 and 6.1, 3.5 and 3.2, and 5.6 and 6.2%, respectively, by ICP/MS, FAAS, and ISE. The fewest matrix effects were found with ICP/MS, followed by FAAS, and ISE. FAAS gave higher results in a variety of matrixes when compared to ICP/MS and/or ISE. ISE did not perform well in fatty foods or at very low Na concentrations. Manufacturers' Nutrition Facts Panel sodium declarations exceeded levels found by analysis in > 70% of the foods. Analysis of chloride content does not produce reliable Na estimates in low-sodium foods, even when added sodium chloride is present. Methodological issues and contamination sources are discussed.