An Exploratory Study Of Estimation Of 24-Hour Urinary Sodium And Potassium Excretion Using Spot Urine Among Cardiovascular And Cerebrovascular Diseases In Rural Liaoning

Objective:To establish an equation to estimate 24-hour urinary sodium(24-h UNa)and potassium(24-h UK)values by spot urine concentrations in patients with cardiovascular and cerebrovascular diseases in rural Liaoning and assess the accuracy of the equations,and assess the applicability and accuracy of published equations.Methods:This study was a sub-study of the Salt Substitute and Stroke Study(SSa SS).The subjects were from 120 villages in 2 counties of rural communities in Liaoning Province.Each year,6 villages were randomly selected by county stratification,and 20 subjects were randomly selected from each village to collect the casual spot urine samples and 24-h urine samples.There were five process index surveys in the whole study.Due to the influence of gender on urinary sodium and potassium metabolism,the model was constructed according to different genders.Sex-specific equation was developed using multiple linear regression analysis.And adjusted R-Square to evaluate the fitting performance,Bootstrap self-sampling was used for internal verification.Intraclass correlation coefficient(ICC)and Bland-Altman analysis were used for the assessment of consistency of the equation between the estimated and measured 24-h UNa and 24-h UK excretion values.Meanwhile differences at the individual level were evaluated by deviation and misclassification analysis in 24-h UNa excretion.Results:A total of 466 participants were included in the study.The median age of the participants was 64.00(9.25)years,82.62%a history of hypertension,and 87.77%had a history of stroke.The median 24-h UNa and 24-h UK excretions at the population level were 3678.85(2567.20)mg/d and 2042.51(1412.60)mg/d,respectively.The median estimates from the developed equation of 24-h UNa and 24-h UK excretion were3858.42(2676.23)mg/d and 1992.95(1440.85)mg/d,respectively.The correlation coefficients of the New equation of 24-h UNa versus 24-h UK were 0.784 and 0.646.In the published formulas,except Kawasaki formula,24-h UNa excretions was underestimated.There was a low to moderate correlation between the estimated values and the measured values,with ICC coefficients less than 0.4,indicating poor reliability.Published formulas all underestimated 24-h UK excretion,and Spearman rank-correlation analysis showed a moderate correlation(r_s>0.5).Bland-Altman analysis showed that most of the points were scattered within the 95%CI.At the individual level of 24-h UNa excretion,the proportions of relative deviations within±10%were 16.95%(Kawasaki),18.03%(Tanaka),14.59%(INTERSALT1),14.59%(INTERSALT1),and 29.40%(New),respectively.The misclassification rate of the published formulas is more than 60%,and the new equations was 43.56%.Conclusions:The new equations exhibited better performance in the study.The methods had the potential to predict individual level 24-h UNa excretion.