Kong Ji-Sook, Lee Yeon-Kyung, Kim Mi Kyung, Choi Mi-Kyeong, Heo Young-Ran, Hyun Taisun, Kim Sun Mee, Lyu Eun-Soon, Oh Se-Young, Park Hae-Ryun, Rhee Moo-Yong, Ro Hee-Kyong, Song Mi Kyung
Division of Cancer Epidemiology and Prevention, National Cancer Center, Goyang, Gyeonggi, South Korea.
Department of Food Science and Nutrition, Kyungpook National University, Daegu, South Korea.
PLoS One. 2018 Feb 15;13(2):e0192588. doi: 10.1371/journal.pone.0192588. eCollection 2018.
This study was conducted to develop an equation for estimation of 24-h urinary-sodium excretion that can serve as an alternative to 24-h dietary recall and 24-h urine collection for normotensive Korean adults. In total, data on 640 healthy Korean adults aged 19 to 69 years from 4 regions of the country were collected as a training set. In order to externally validate the equation developed from that training set, 200 subjects were recruited independently as a validation set. Due to heterogeneity by gender, we constructed a gender-specific equation for estimation of 24-h urinary-sodium excretion by using a multivariable linear regression model and assessed the performance of the developed equation in validation set. The best model consisted of age, body weight, dietary behavior ('eating salty food', 'Kimchi consumption', 'Korean soup or stew consumption', 'soy sauce or red pepper paste consumption'), and smoking status in men, and age, body weight, dietary behavior ('salt preference', 'eating salty food', 'checking sodium content for processed foods', 'nut consumption'), and smoking status in women, respectively. When this model was tested in the external validation set, the mean bias between the measured and estimated 24-h urinary-sodium excretion from Bland-Altman plots was -1.92 (95% CI: -113, 110) mmol/d for men and -1.51 (95% CI: -90.6, 87.6) mmol/d for women. The cut-points of sodium intake calculated based on the equations were ≥4,000 mg/d for men and ≥3,500 mg/d for women, with 89.8 and 76.6% sensitivity and 29.3 and 64.2% specificity, respectively. In this study, a habitual 24-hour urinary-sodium-excretion-estimation model of normotensive Korean adults based on anthropometric and lifestyle factors was developed and showed feasibility for an asymptomatic population.
本研究旨在开发一种估算24小时尿钠排泄量的方程，该方程可作为正常血压韩国成年人24小时饮食回顾和24小时尿液收集的替代方法。总共收集了来自该国4个地区的640名年龄在19至69岁之间的健康韩国成年人的数据作为训练集。为了对从该训练集开发的方程进行外部验证，独立招募了200名受试者作为验证集。由于性别差异，我们使用多变量线性回归模型构建了一个性别特异性方程来估算24小时尿钠排泄量，并评估了所开发方程在验证集中的性能。最佳模型在男性中分别由年龄、体重、饮食行为（“吃咸食”、“食用泡菜”、“食用韩式汤或炖菜”、“食用酱油或红辣椒酱”）和吸烟状况组成，在女性中分别由年龄、体重、饮食行为（“盐偏好”、“吃咸食”、“检查加工食品的钠含量”、“食用坚果”）和吸烟状况组成。当在外部验证集中测试该模型时，Bland-Altman图中测量的和估计的24小时尿钠排泄量之间的平均偏差在男性中为-1.92（95%CI：-113，110）mmol/d，在女性中为-1.51（95%CI：-90.6，87.6）mmol/d。根据方程计算的钠摄入量切点在男性中≥4000mg/d，在女性中≥3500mg/d，敏感性分别为89.8%和76.6%，特异性分别为29.3%和64.2%。在本研究中，基于人体测量和生活方式因素开发了正常血压韩国成年人习惯性24小时尿钠排泄量估算模型，并显示了对无症状人群的可行性。
