The accuracy of estimating calcium, sodium and potassium intake with a food record

Abstract

The underreporting of energy intake by food recording is well known, but less is known about accuracy of micronutrient assessment by food records. The purpose of this study was to determine whether the underreporting of calcium, sodium and potassium intake is associated with energy underreporting as assessed by food diary. The subjects were 94 healthy, non-smoking premenopausal women. They kept a 3-day food diary. The mean basal metabolic rate (BMR) was estimated by the equation of FAO/WHO/UNU. The ratio of BMR to reported energy intake (EI) was calculated, the selected limit for underreporting being 1.27. During the last day of the recording period the participants collected a 24-hour urine sample by which the electrolyte excretion was assessed and used as the criterion for the accuracy of the food recording. The ratio between excretion and intake (EIR) was calculated as the percentage of the intake. The average daily intakes of all the nutrients were lower for the low energy reporters (LER≤1.27 * BMR) than for the control reporters (CR>1.27 * BMR). The total median EI was 6430 kJ/d compared with 8989 kJ/d (P<0.001) (P for group differences). The median daily intake was 27.9 and 30.5 mmol (P=0.12) for calcium, 112.9 and 142.1 mmol (P<0.001) for sodium, and 75.6 and 90.2 mmol (P<0.001) for potassium for the LERs and the CRs, respectively. The median calcium/creatinine excretion values were 0.42 and 0.40, sodium/creatinine 11.0 and 11.6, and potassium/creatinine 5.8 and 6.5 for LERs and CRs, respectively. Even though the intake values were significantly lower for sodium and potassium and somewhat lower for calcium in the group of LERs compared to CRs, neither the excretion nor EIR values differed. This gives reason to suspect that low energy and nutrient intake of the LERs was partly due to underreported food intake but also due to decreased eating during the recording period.