Mercado Carla I, Cogswell Mary E, Valderrama Amy L, Wang Chia-Yih, Loria Catherine M, Moshfegh Alanna J, Rhodes Donna G, Carriquiry Alicia L
From the Division for Heart Disease and Stroke Prevention, National Center for Chronic Disease Prevention and Health Promotion, Atlanta, GA (CIM, MEC, and ALV); the Division of Health and Nutrition Examination Surveys, National Center for Health Statistics, Hyattsville, MD (C-YW), the Centers for Disease Control and Prevention; National Heart, Lung, and Blood Institute, NIH, Bethesda, MD (CML); the Beltsville Human Nutrition Research Center, Agricultural Research Service, USDA, Beltsville, MD (AJM and DGR); and the Department of Statistics, Iowa State University, Ames, IA (ALC).
Am J Clin Nutr. 2015 Feb;101(2):376-86. doi: 10.3945/ajcn.113.081604. Epub 2014 Dec 17.
Limited data are available on the accuracy of 24-h dietary recalls used to monitor US sodium and potassium intakes.
We examined the difference in usual sodium and potassium intakes estimated from 24-h dietary recalls and urine collections.
We used data from a cross-sectional study in 402 participants aged 18-39 y (∼50% African American) in the Washington, DC, metropolitan area in 2011. We estimated means and percentiles of usual intakes of daily dietary sodium (dNa) and potassium (dK) and 24-h urine excretion of sodium (uNa) and potassium (uK). We examined Spearman's correlations and differences between estimates from dietary and urine measures. Multiple linear regressions were used to evaluate the factors associated with the difference between dietary and urine measures.
Mean differences between diet and urine estimates were higher in men [dNa - uNa (95% CI) = 936.8 (787.1, 1086.5) mg/d and dK - uK = 571.3 (448.3, 694.3) mg/d] than in women [dNa - uNa (95% CI) = 108.3 (11.1, 205.4) mg/d and dK - uK = 163.4 (85.3, 241.5 mg/d)]. Percentile distributions of diet and urine estimates for sodium and potassium differed for men. Spearman's correlations between measures were 0.16 for men and 0.25 for women for sodium and 0.39 for men and 0.29 for women for potassium. Urinary creatinine, total caloric intake, and percentages of nutrient intake from mixed dishes were independently and consistently associated with the differences between diet and urine estimates of sodium and potassium intake. For men, body mass index was also associated. Race was associated with differences in estimates of potassium intake.
Low correlations and differences between dietary and urinary sodium or potassium may be due to measurement error in one or both estimates. Future analyses using these methods to assess sodium and potassium intake in relation to health outcomes may consider stratifying by factors associated with the differences in estimates from these methods. This trial was registered at clinicaltrials.gov as NCT01631240.
关于用于监测美国人群钠和钾摄入量的24小时膳食回顾法的准确性，现有数据有限。
我们研究了通过24小时膳食回顾法和尿液收集法估算的日常钠和钾摄入量之间的差异。
我们使用了2011年在华盛顿特区大都市区进行的一项横断面研究的数据，该研究涉及402名年龄在18 - 39岁（约50%为非裔美国人）的参与者。我们估算了日常膳食钠（dNa）和钾（dK）摄入量的均值和百分位数，以及24小时尿钠（uNa）和尿钾（uK）排泄量。我们研究了膳食测量法和尿液测量法估算值之间的斯皮尔曼相关性及差异。使用多元线性回归评估与膳食和尿液测量法差异相关的因素。
男性膳食和尿液估算值之间的平均差异高于女性 [dNa - uNa（95%置信区间）= 936.8（787.1，1086.5）mg/d，dK - uK = 571.3（448.3，694.3）mg/d] [女性：dNa - uNa（95%置信区间）= 108.3（11.1，205.4）mg/d，dK - uK = 163.4（85.3，241.5）mg/d]。男性钠和钾的膳食和尿液估算值的百分位数分布有所不同。钠测量值之间的斯皮尔曼相关性男性为0.16，女性为0.25；钾测量值之间的斯皮尔曼相关性男性为0.39，女性为0.29。尿肌酐、总热量摄入以及混合菜肴中营养素摄入量的百分比与钠和钾摄入量的膳食和尿液估算值之间的差异独立且持续相关。对于男性，体重指数也与之相关。种族与钾摄入量估算值的差异相关。
膳食和尿钠或尿钾之间的低相关性和差异可能是由于一种或两种估算方法存在测量误差。未来使用这些方法评估钠和钾摄入量与健康结局之间关系的分析可能需要考虑按与这些方法估算值差异相关的因素进行分层。该试验已在clinicaltrials.gov上注册，注册号为NCT01631240。
