Reporting accuracy of population dietary sodium intake using duplicate 24 h dietary recalls and a salt questionnaire

Abstract

High dietary Na intake is associated with multiple health risks, making accurate assessment of population dietary Na intake critical. In the present study, reporting accuracy of dietary Na intake was evaluated by 24 h urinary Na excretion using the EPIC-Soft 24 h dietary recall (24-HDR). Participants from a subsample of the European Food Consumption Validation study (n 365; countries: Belgium, Norway and Czech Republic), aged 45-65 years, completed two 24 h urine collections and two 24-HDR. Reporting accuracy was calculated as the ratio of reported Na intake to that estimated from the urinary biomarker. A questionnaire on salt use was completed in order to assess the discretionary use of table and cooking salt. The reporting accuracy of dietary Na intake was assessed using two scenarios: (1) a salt adjustment procedure using data from the salt questionnaire; (2) without salt adjustment. Overall, reporting accuracy improved when data from the salt questionnaire were included. The mean reporting accuracy was 0·67 (95 % CI 0·62, 0·72), 0·73 (95 % CI 0·68, 0·79) and 0·79 (95 % CI 0·74, 0·85) for Belgium, Norway and Czech Republic, respectively. Reporting accuracy decreased with increasing BMI among male subjects in all the three countries. For women from Belgium and Norway, reporting accuracy was highest among those classified as obese (BMI $30 kg/m 2 : 0·73, 95 % CI 0·67, 0·81 and 0·81, 95 % CI 0·77, 0·86, respectively). The findings from the present study showed considerable underestimation of dietary Na intake assessed using two 24-HDR. The questionnaire-based salt adjustment procedure improved reporting accuracy by 7 -13 %. Further development of both the questionnaire and EPIC-Soft databases (e.g. inclusion of a facet to describe salt content) is necessary to estimate population dietary Na intakes accurately.Key words: Diet surveys: Self-reports: Biological markers/urine: Dietary sodium: European Food Consumption ValidationThe current WHO guidelines strongly recommend a reduction in Na intake to ,2 g/d (i.e. , 5 g salt/d) (1) . There is conclusive evidence that reduction in Na intake reduces blood pressure (2 -5) . The relationship between Na consumption and the risk of CVD and stroke is less clear than that of hypertension. For example, a systematic review of randomised controlled trials has found no relationship between Na intake and CVD risk (6) ; however, a meta-analysis of thirteen cohort studies has concluded that there is a direct relationship between increased Na consumption and the subsequent risk of CVD and stroke (7) . Data from an international epidemiological study on dietary factors in the aetiology of unfavourable blood pressure (the INTERMAP study) showed that sources of dietary Na can vary considerably. For instance, processed foods contribute heavily to Na intake in the USA and the UK (70-95 %), whereas in China, most dietary Na intake (76 %) is from salt added during home cooking (8) . Another study, in which Li

Twenty‐Four–Hour Diet recall and Diet records compared with 24‐hour urinary excretion to predict an individual’s sodium consumption: A Systematic Review

“…In the pooled analysis by Freedman et al, the attenuation factors for 24‐hour diet recalls were higher than those calculated for Food Frequency Questionnaires in the same studies, indicating that estimates from 24‐hour diet recalls are more suitable than Food Frequency Questionnaires for use in cohort studies investigating relationships between sodium intake and disease outcomes, because there is less attenuation of the relationship due to measurement error of sodium intake . Other methods of comparison included reporting a ratio of the means from two methods (dietary estimate/urinary sodium excretion) to estimate reporting accuracy …”

“…16 Other methods of comparison included reporting a ratio of the means from two methods (dietary estimate/urinary sodium excretion) to estimate reporting accuracy. 19,31,37,42,43 As with our first review evaluating Food Frequency Questionnaires, 15 there was variability between studies as to number of days of dietary assessment, number of 24-hour urines collected, and interpretation of 24-hour urine results. Many of the 24-hour diet recall studies reported using multiple-pass methods of assessment, a technique that has been developed to minimize recall bias which involves an initial overview of foods consumed in the period, followed by more detailed questioning of each eating occasion.…”

Perspective: The Application of A Priori Diet Quality Scores to Cardiovascular Disease Risk—A Critical Evaluation of Current Scoring Systems

“…Basing the DASH diet score on an FFQ raises concerns, especially for quantifying sodium intake. For large-scale cohort studies, an alternative approach for measuring sodium intake used a salt-based questionnaire that improved accuracy by 7–13% ( 99 ).…”

Dietary reference values for sodium

“…For samples with a PABA recovery between 50% and 85%, urinary Na excretion was corrected (formula from Johansson et al. ()) 45–6560173N/A161–185N/AN/ACzech Republic6018256 173 P5‐P95 101–288 Norway (De Keyzer et al., ) 62 Subsamples of the European EFCOVAL study; subjects recruited by convenience sampling through advertisements 151N/A141–161N/AN/A Croatia (Premužić et al., ) 2009 24‐h urine collection; completeness check not reported 46.3 ± 7.3 N/A (total sample of 93 men and women); From two out‐patient clinics (one urban, one rural); ‘salt‐mapping survey’ 17773N/AN/A Croatia (Dika et al., ) 2009Morning spot urine samples; 24‐h Na excretion was estimated by applying Kawasaki, INTERSALT and Tanaka equationsN/A N/A (total sample of 1,669 men and women); Random sample (door‐to‐door method) in the continental rural part of Croatia Kawasaki: 21478N/AN/A INTERSALT: 22256N/AN/A Tanaka: 17869N/AN/A Finland (Laatikainen et al., )2002 24‐h urine collection; exclusion of incomplete samples with Cr levels ≤ 5.0 mmol/day or Cr levels ≤ 6.0 mmol/day together with a urine volume < 1,000 mL 25–64…”