Yu Dahai, Cai Yamei, Qin Rui, Tian Xing, Xiao Jing, Zhao Zhanzheng
Department of Nephrology, the First Affiliated Hospital, Zhengzhou University, Zhengzhou, China.
Arthritis Research UK Primary Care Centre, Research Institute for Primary Care & Health Sciences, Keele University, Keele, United Kingdom.
Kidney Blood Press Res. 2018;43(1):162-169. doi: 10.1159/000487498. Epub 2018 Feb 16.
BACKGROUND/AIMS: During peritoneal equilibrium test (PET), intermediate measures of Dt/P (dialysate/plasma creatinine ratio), D0/P, D2/P and D4/P ratios might provide additional information regarding a patient's cardiovascular mortality. We aimed to characterise heterogeneity of dialysate creatinine response patterns based on three ratios and compare cardiovascular mortality risks by response patterns.
3,477 patients initialising peritoneal dialysis (PD) between 2007-2015 were measured D0/P, D2/P and D4/P at baseline and 2-year cardiovascular mortality was defined as the primary outcome. Latent class mixed-effect models were fitted to identify distinct patterns of dialysate creatinine response. Multivariable Logistic regression model with adjustment of cardiometabolic factors were used to compare cardiovascular mortality by latent classes.
The 1st latent class including 36% of participants was characterised by consistently increasing and high Dt/P ratios both at 2-hour and 4-hour. The 2nd class including 61% of participants characterised by sharply increased ratio at 2-hour and slightly increased ratio at 4-hour. The 3rd class included 3% of participants with decreased ratio at 2-hour and increased ratio at 4-hour. Comparing the 1st class, the lower cardiovascular mortality risk was identified in the 2nd and 3rd class with adjusted odds ratio 0.65 (95% confidence interval: 0.62-0.69) and 0.48 (0.41-0.57), respectively.
Patients with decreased Dt/P ratio between 0-hour and 2-hour and low ratios at 2-hour and 4-hour tend to have low cardiovascular mortality. Latent class analysis seems to be a promising method to reveal unidentified subgroups that do not fit into the risk category defined by Dt/P ratio at 4-hour.
背景/目的：在腹膜平衡试验（PET）期间，Dt/P（透析液/血浆肌酐比值）、D0/P、D2/P和D4/P比值的中间测量值可能会提供有关患者心血管死亡率的额外信息。我们旨在根据三个比值来描述透析液肌酐反应模式的异质性，并比较不同反应模式的心血管死亡风险。
对2007年至2015年间开始进行腹膜透析（PD）的3477例患者在基线时测量D0/P、D2/P和D4/P，并将2年心血管死亡率定义为主要结局。采用潜在类别混合效应模型来识别透析液肌酐反应的不同模式。使用调整了心脏代谢因素的多变量逻辑回归模型来比较不同潜在类别的心血管死亡率。
第一个潜在类别包括36%的参与者，其特征是在2小时和4小时时Dt/P比值持续升高且较高。第二个类别包括61%的参与者，其特征是2小时时比值急剧升高，4小时时比值略有升高。第三个类别包括3%的参与者，其特征是2小时时比值降低，4小时时比值升高。与第一个类别相比，第二个和第三个类别的心血管死亡风险较低，调整后的优势比分别为0.65（95%置信区间：0.62-0.69）和0.48（0.41-0.57）。
0小时至2小时之间Dt/P比值降低且2小时和4小时时比值较低的患者心血管死亡率往往较低。潜在类别分析似乎是一种有前景的方法，可揭示不符合4小时时Dt/P比值所定义风险类别的未识别亚组。
