Validation of salt intake measurements: comparisons of a food record checklist and spot-urine collection to 24-h urine collection.

Abstract

Monitoring population salt intake is operationally and economically challenging. We explored whether a questionnaire assessment and a prediction of Na intake from spot-urine could replace or complement the recommended measurement of Na in 24-h urine (24-h U). Compare the agreement of a Na-specific food record checklist (FRCL) and a late-afternoon spot-urine measurement (PM-spot) with 24-h U measurement in estimating Na intake at group level. Each participant's use of these methods extended over 3 d. Agreement was assessed using mean (95 % CI) differences, linear regression models and Bland-Altman plots. The validation study was part of a 1-year workplace intervention trial to lower salt intake in Switzerland. Seventy women and 71 men, aged 21-61 years, completed three FRCL, and acceptable PM-spot and 24-h U samples at baseline (April-October 2015). Mean Na intake estimates varied slightly across methods (3·5-3·9 g/d). Mean Na intake differences from 24-h U were 0·2 (95 % CI (0, 0·5)) g/d for FRCL and 0·4 (95 % CI (0·2, 0·6)) g/d for PM-spot. Linear regression models and Bland-Altmann plots more clearly depicted differences by sex and discretionary salt use. Although 24-h U remains the best reference method for monitoring Na intake at the population level, PM-spot and FRCL might be more practical instruments for frequent, periodic Na intake assessments. Population-specific prediction models to estimate 24-h U could be developed and evaluated.