Feasibility of using spot urines as surrogates for 24‐hour urines to assess sodium and potassium excretion for CVD association studies

Abstract

IntroductionAssessing the association of dietary sodium and potassium with CVD risk is hampered by biases in dietary self‐report and the participant burden and cost of obtaining 24‐hour urine collections for biomarker analysis. Sodium and potassium levels measured from 24‐hour urines are considered reliable biomarkers of corresponding dietary intake. Spot urines, which are more practical for epidemiologic research, have been utilized, but the measurement properties are uncertain.ObjectiveTo elucidate the measurement properties of spot urine‐derived sodium and potassium excretion estimates through comparison with 24‐hour urine for sodium and potassium biomarkers.Research DesignWe conducted a feasibility study within the Women's Health Initiative (WHI) ongoing biomarker sub‐studies to assess the association between spot and 24‐hour urinary sodium and potassium. The WHI is a long‐term study (1993‐present) among postmenopausal women across the United States. Sodium and potassium were analyzed by ion selective electrode and creatinine by spectrophotometric detection of a colored creatinine‐picrate complex. Creatinine was used as an adjustment factor linking the spot and 24‐hour sodium and potassium concentrations. Statistical analyses generated calibration equations for estimating 24‐hour sodium and potassium excretion from spot urines and utilizing study subject characteristics.ResultsBoth a spot urine and 24‐hour urine were collected from n=150 participants wherein sodium, potassium and creatinine were assayed (Table 1).Calibration equations for creatinine‐adjusted sodium, potassium, and sodium/potassium ratio were developed separately by regressing 24‐hour excretion levels on spot urine measurements and subject characteristics selected among age, body mass index, physical activity, education, season, and race, on the basis of Akaike information criterion (AIC) (Table 2).The ability to estimate 24‐hour excretions from spot urine was similar whether or not these measurements were log‐transformed. Results do not suggest heterogeneity in variance of spot urine excretion levels or calibration estimates as a function of 24 hour excretion level.ConclusionsCorrelations (square root of R2 values) in the vicinity of 0.4–0.5 after adjusting for creatinine suggest that spot urine may serve as a suitable replacement for 24‐hour urine for assessing sodium and potassium intake. This feasibility study shows that WHI has the potential to cost‐effectively use existing biospecimens within WHI sub‐cohort of approximately 11,020 postmenopausal women with spot urines to research the association of sodium and potassium intakes as risk factors for CVD in a diverse sample of US postmenopausal women.Support or Funding InformationNational Cancer Institute (NCI) grant R01 CA119171 National Heart, Lung, and Blood Institute, National Institutes of Health, U.S. Department of Health and Human Services through contracts, HHSN268201100046C, HHSN268201100001C, HHSN268201100002C, HHSN268201100003C, HHSN268201100004C Urinary excretion of sodium and potassium and ratio by creatinine (geometric means) in spot and 24‐hour collections (n=150) Sodium (Na) Potassium (K) Spot 24‐hour Spot 24‐hour Excretion, mEq/L 59.5 45.2 27.5 31.3 Excretion ratio Na or K/creatinine* 1.10 1.04 0.51 0.70 Na or K (mEq/L)/creatinine (mg/dL) Percentage of variance (R2) accounted for upon regressing 24‐hour urinary excretion levels of sodium and potassium (log‐transformed) on spot urine measurements* Sodium (Na) Potassium (K) Na/K R2 without subject characteristics 19.6% 26.7% 18.0% R2 with subject characteristics** 23.3% 31.1% 19.7% Creatinine‐adjusted Age, body mass index, physical activity, education, season, and race