Dietary sodium estimation methods: accuracy and limitations of old and new methods in individuals at high cardiovascular risk.

Christiana Tsirimiagkou,Theodore G Papaioannou,Antonios A Argyris,Eirini D Basdeki,Kalliopi Karatzi,Athanase D Protogerou,Maria Yannakoulia

doi:10.1017/s1368980021004390

Abstract

Accurate and easy to use methods for dietary Na intake estimation in population level are lacking. We aimed at (i) estimating the mean Na intake in the group level using a variety of dietary methods (DM) and urinary methods (UM) and correlating them with 24-h urine collection (24UCol) and (ii) improving the accuracy of the existing DM. The most common DM (three 24-h dietary recalls (24DR) and FFQ) and UM (24UCol and spot urine collection using common equations) were applied. To improve the existing: (i) 24DR, discretionary Na was quantified using salt-related questions or adding extra 15 % in total Na intake and (ii) FFQ, food items rich in Na and salt-related questions were added in the standard questionnaire (NaFFQ). National and Kapodistrian University of Athens, Greece. Totally, 122 high cardiovascular risk subjects (56·0 ± 12·6 years; 55·7 % males). Mean 24 h Na excretion (24UNa) was 2810 ± 1304 mg/d. Spot urine methods overestimated the 24UNa (bias range: -1781 to -492 mg) and were moderately correlated to 24UCol (r = 0·469-0·596, P ≤ 0·01). DM underestimated the 24UNa (bias range: 877 to 1212 mg) and were weakly correlated with 24UCol. The improved DM underestimated the 24UNa (bias range: 877 to 923 mg). The NaFFQ presented the smallest bias (-290 ± 1336 mg) and the strongest correlation with 24UCol (r = 0·497, P ≤ 0·01), but wide limits of agreement in Bland-Altman plots (-2909 mg; 2329 mg), like all the other methods did. The existing methods exhibit poor accuracy. Further improvement of the newly developed NaFFQ could be promising for more accurate estimation of mean dietary Na intake in epidemiological studies. Additional validation studies are needed.

Highlights

In order to detect a minimum difference of 500 mg in daily Na intake between each Na estimation method and the 24-h urine Na excretion (24UNa) (α = 0·05, power = 0·80), the minimum sample size for each pair of methods was calculated (n 60)(27,28)
One hundred and twenty-two [122] participants with available 24-h urine collection (24UCol) data were used for the analyses (56·0 ± 12·6 years; 55·7 % males) (Table 2)
The INTERSALT equation presented the lowest bias among all the other equations; it was moderately correlated with 24UNa and presented consistent bias in

Summary

Introduction

The main findings of this study are (i) the existing DM tend to underestimate and spot urine methods tend to overestimate the true Na intake; (ii) all the existing DM are weakly correlated and present poor agreement with the 24UCol, and all the spot urine methods are moderately correlated and present moderate agreement with the 24UCol and (iii) the new NaFFQ is the only method that performed better in the analysis, having simultaneously the smallest bias in mean differences, the strongest correlation with the 24UCol regarding DM and a homogeneous variation as the mean Na intake increases in Bland–Altman plots, but still wide limits of agreement.

Results

Conclusion