Single 24-hour urine collection does not suffice to estimate an individual’s sodium intake
Use of a Single Baseline Versus Multi-Year 24-Hour Urine Collections for Estimation of Long-Term Sodium Intake and Associated Cardiovascular and Renal RiskOlde Engberink RHG, van den Hoek TC, van Noordenne ND, et al., - Circulation. 2017; Originally published June 27, 2017. https://doi.org/10.1161/CIRCULATIONAHA.117.029028
It is well-established that high sodium intake increases blood pressure (BP). Since BP lowering is known to reduce CV events and kidney disease progression, it is expected that the BP lowering via a reduction in sodium intake will have CV and renal protective effects. Recent prospective cohort studies show, however, inconsistent results about the optimal sodium intake for CV and renal protection.
No adequately powered randomized controlled studies have assessed the effects of sodium reduction on CV outcomes. Observations from prospective cohort studies should be interpreted with caution, as reverse causality, residual confounding and incorrect estimation of dietary sodium intake may affect the observed associations [1,2].
Most cohort studies rely on a single collection of morning or 24-hour urine sample to estimate long-term sodium intake. While a single 24-hour urine sample is generally considered the best way to assess sodium intake in clinical practice, it may be an inaccurate representation of long-term sodium intake, due to day-to-day variation in intake, susceptibility to collection errors and potential influence of medication. Moreover, 24-hour sodium excretion has been described to differ from 24-hour sodium intake during fixed sodium intake [3,4].
This retrospective cohort study compared the value of using a single baseline 24-hour samples with using at least 1 follow-up samples for estimation of sodium intake, as well as the impact of sodium intake on long-term outcome in subjects with preserved kidney function (eGFR >60 ml/1.73m2/min). 574 Subjects were eligible and 401 had at least one follow-up measurement in the first year following inclusion and 529 had an additional measurement within 5 years. Multiple measurements in a person were averaged per time period. During a median follow-up of 12.7 years, 9776 24-hour urine collections were analyzed.
- Average 24-hour sodium excretion at baseline was 3.9±1.0 grams (169±74 mmol).
- Average 24-hour urine sodium excretion showed similar results between a single baseline measurement and follow-up measurements at 1, 5 or 15 years (P=0.88).
- Significant intra-individual changes in 24-hour urinary sodium excretion were observed over time. In most subjects, estimated sodium intake using a single baseline measurement differed at least 0.4 gram sodium (17 mmol sodium, 1 gram of salt) from the 1-year (71%), 5-year (72%) or 15-year (75%) results. In 49%, 50% and 52%, respectively, of the subjects the difference was even 0.8 grams of sodium.
- Tertile classification changed in 45%, 49% and 50% of the 1-year, 5-year and 15-year measurements respectively, as compared with a single baseline measurement.
- Based on a single measurement for sodium intake estimation, high 24-hour sodium excretion was not associated with an increased risk for the composite of CV events and mortality, when compared with low sodium intake (HR: 1.09, 95%CI: 0.61-1.95).
- In 24-hour sodium excretion measurements obtained within 1 year or within 5 years after baseline, high sodium excretion was associated with a higher risk for CV events and death (1-year: HR: 1.80, 95%CI: 1.03-3.13, 5-years HR: 1.73, 95%CI: 1.00-2.99). Hazards for CV events and death of long-term estimates were up to 85% different from baseline estimates.
- Long-term estimates of hazard for renal outcome were up to 47% different from baseline single 24-hour urine excretion data. Inconsistencies were even larger when data were analyzed separately for those with and those without primary kidney disease.
This study shows that, on a population level, a single baseline measurement and repetitive follow-up 24-hour urine sodium measurements lead to similar estimates of sodium intake. Estimations of individual sodium intake, however, differ significantly. This affected the association between sodium intake and long-term outcomes.
These data indicate that a single 24-hour urine sodium collection can be used to estimate average sodium intake in a population, but is imprecise for estimation of long-term individual sodium intake, and thus inadequate to study the relation between sodium intake and long-term CV or renal outcome.
Imprecise estimation of sodium intake at baseline has consequences for classification of subjects as low, moderate or high sodium consumers. Only when 1- and 5-year estimates were used, a positive correlation between sodium intake and long term CV risks was seen.