Validation of a food frequency questionnaire and a spot urine sample for the assessment of dietary sodium intake in Lebanese adults -

Abstract

Non-communicable diseases (NCDs) are globally on the rise. In 2010, cardiovascular diseases (CVD) were reported as the primary cause of disability and death in the Arab world. Diet and lifestyle play an important role in determining the risk of CVDs and evidence suggests a link between salt intake and CVDs which sheds the light on the importance of assessing the levels of population salt intake. In Lebanon, the scarcity of data on dietary sodium intake highlights the need for rigorous investigations aiming at assessing the population’s intake of salt and sodium. The gold-standard for the assessment of sodium intake is 24 hour urinary sodium excretion. Even though alternative methods have been proposed for the dietary assessment of Na, the validity, accuracy and-or applicability of these methods have been often criticized in the literature. Objectives: The objective of the present study is to validate 2 different methods for the assessment of Na intake in adults, namely the sport urinary sodium excretion and the food-frequency questionnaire. Validation will be undertaken through the assessment of sodium urinary excretion, which is recognized as the gold standard in the assessment of sodium intakes. Methods: A convenience sample of Lebanese adults aged 19 to 55 years were recruited. Subjects provided a 24 hour urine collection and a spot urine sample (particularly the second morning voiding). Subjects were also asked to complete an FFQ on the day of submission of urine samples. Anthropometric measurements (weight, and height) were obtained. Descriptive statistics, Spearman correlations, mean differences, and Bland-Altman plots, Intraclass Correlation (ICC), weighted Kappa, and percent agreement were used to assess the validity of the FFQ and two spot urine formulas: Kawasaki and Tanaka against the gold standard in assessing sodium intake, 24 hour urine. SPSS was used in the analysis and a p0.05 indicated significance. Results: Out of the 72 subjects, 60 completed the study. Spearman correlation coeffic