Combination of a modified dietary tool and spot urine collection: an improved method for daily sodium intake estimation.

PURPOSE: Recently, an improved food frequency questionnaire focused on assessing daily sodium (Na) intake (Food Frequency Questionnaire [FFQ]-Na: 73 questions; time to be filled in [t]: 25 min) was developed with higher accuracy over other dietary and spot urinary methods (sUMs) compared with the gold standard 24 h urinary Na (24hUNa) excretion, which approaches real daily Na intake. We aimed to improve the FFQ-Na in terms of usability and combine it with sUMs to test whether daily Na intake assessment could be further improved for use in clinical practice and not only in research. METHODS: In a cross-sectional study, 219 participants (61.2 males, age 55.45 ± 12.24 y) were asked to fill in the FFQ-Na and provide two 24hUNa collections and one spot urine sample. Data were analyzed using artificial intelligence and receiver operating characteristic curve analysis. RESULTS: Using artificial intelligence: 1) Three new shorter versions of the FFQ-Na were developed by discarding questions with low additive value: FFQ-Na-1 (43 questions; t = 15 min), FFQ-Na-2 (37 questions; t = 13 min) and FFQ-Na-3 (29 questions; t = 10 min); 2) the FFQ-Na-2 in combination with the Tanaka (r2 = 0.404) or Toft equation (r2 = 0.411) had the best performance compared with 24hUNa, indicating moderate predictive ability. Receiver operating characteristic curve analysis for the detection of daily intake of >3 g Na/d indicated that the FFQ-Na-2 displayed the highest area under the curve (0.721), sensitivity (0.628), and specificity (0.839). CONCLUSION: The FFQ-Na-2, in combination with a sUM (Tanaka or Toft), is a promising method for daily dietary Na intake assessment (using 24hUNa collection, which is the best method available to assess daily Na intake); further improvement and tests are needed.