Abstract
BACKGROUND: Dietary assessment methods such as food frequency questionnaires (FFQs) and 24-hour food records (24hFRs) are widely used in population-based studies due to their feasibility and relatively low cost. However, these self-reported methods are subject to measurement errors, including recall bias and misreporting, which may vary across individuals and seasons. In contrast, 24-hour urinary excretion (24UE) of sodium, potassium, and nitrogen is considered a more objective biomarker of intake and is often used as a reference method in validation studies. AIM: This study aimed to evaluate the agreement of a 24-hour food record (24hFR) and the relative validity of a food frequency questionnaire (FFQ) against recovery urinary biomarkers of sodium, potassium, and urea-derived protein intake. METHOD: This observational, cross-sectional validation study included a total of 37 healthy adult women aged 18–44 years. Information on participants’ sociodemographic characteristics, dietary habits, and physical activity levels was collected, and anthropometric measurements were performed by the researcher. In addition, 24-hour food records (24hFR) and 24-hour urine excretion (24UE) samples were collected twice in summer and twice in winter on nonconsecutive days with a 14-day interval, resulting in a total of four collections per participant, and food consumption frequencies (FFQs) were determined. RESULTS: The most commonly used salt type among the participants was iodized table salt (56.8%). The mean sodium intake estimated from the 24-hour food record was significantly higher in the winter (3142.2 ± 905.7 mg/day) compared to the summer (2479.6 ± 793.5 mg/day) (p < 0.05). The mean sodium and potassium intakes estimated by the 24-hour food record were similar to those determined by urinary sodium and potassium excretion at the group level in both seasons, although correlations at the individual level were weak. The difference between the measurement methods for sodium and potassium values (FFQ, 24hFR, 24UE) in summer and winter was due to the frequency of food consumption. Protein intake estimated by the FFQ and food record methods was comparable; however, protein intake calculated from urinary excretion was higher than the estimated values. CONCLUSION: In conclusion, the 24-hour food record method, which is widely used in nutritional assessment, showed closer agreement with urinary sodium and potassium excretion at the group (mean) level rather than at the individual ranking level, but underestimated protein intake in this study. These findings highlight the need for complementary biochemical markers when assessing protein consumption in population-based dietary surveys. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12937-026-01298-7.