Abstract:

Objective　To construct and validate a predictive model for the occurrence of cognitive impairment in patients taking peritoneal dialysis by nomograms. Methods　This was a prospective study. One hundred and twenty patients taking peritoneal dialysis at Department of Nephrology, Yantai Laiyang Central Hospital from January 2020 to January 2024 were selected as the study objects. According to the Montreal Cognitive Assessment (MoCA) scores, the patients were divided into a cognitively normal group (52 cases) whose MoCA score ≥26 and a cognitive impairment group (68 cases) whose MoCA score < 26. There were 28 males and 24 females in the cognitively normal group; they were (50.19±12.79) years old. There were 36 males and 32 females in the cognitive impairment group; they were (53.87±10.84) years old. The general data and clinical indicators in both groups were collected. Multivariate logistic regression analysis was used to identify the risk factors of cognitive impairment in the patients. Based on the results, a nomogram prediction model was established and validated by a calibration curve. t and χ² tests were utilized. Results　The proportion of the patients with diabetes and the levels of glycosylated hemoglobin (HbA1c), parathyroid hormone (PTH), and serum ferritin (SF) in the cognitively normal group were lower than those in the cognitive impairment group [3.85% (2/52) vs. 25.00% (17/68), (5.06±0.31)% vs. (5.52±0.36)%, (304.91±87.79) μg/L vs. (395.69±108.71) μg/L, and (342.83±85.76) μg/L vs. (433.71±134.93) μg/L], with statistical differences (χ²=9.895; t=7.359, 4.918, and 4.243; all P<0.05). The multivariate logistic regression analysis showed that having diabetes and high levels of HbA1c, PTH, and SF were independent risk factors for the occurrence of cognitive impairment in the patients (all P<0.05). Based on these findings, a nomogram prediction model was developed, and it could effectively predict the occurrence of cognitive impairment in the patients. The calibration curve confirmed that the model had good predictive accuracy. Conclusion　Patients with diabetes and high HbA1c, PTH, and SF levels have higher risk of cognitive impairment. These factors should be given special attention in treatment, and appropriate preventive measures should be taken.

Key words:

Peritoneal dialysis, Cognitive impairment, Risk factors, Logistic regression analysis

摘要：

目的　本研究旨在用列线图构建腹膜透析患者认知功能损害发生的预测模型并验证。方法　前瞻性研究，选取2020年1月至2024年1月在烟台市莱阳中心医院肾内科收治的120例腹膜透析患者作为研究对象，根据蒙特利尔认知评估量表（MoCA）评分分为认知正常组52例（MoCA评分≥26分）和认知损害组68例（MoCA<26分）。认知正常组男28例、女24例，年龄（50.19±12.79）岁；认知损害组男36例、女32例，年龄（53.87±10.84）岁。收集两组患者一般资料、临床指标，通过多因素logistic回归分析腹膜透析患者认知功能损害的危险因素，根据结果建立列线图预测模型，并采用校准曲线进行验证。采用t检验、χ²检验。结果　认知正常组患者糖尿病占比及糖化血红蛋白（HbA1c）、甲状旁腺激素（PTH）、铁蛋白（SF）水平均低于认知损害组［3.85%（2/52）比25.00%（17/68）、（5.06±0.31）%比（5.52±0.36）%、（304.91±87.79）μg/L比（395.69±108.71）μg/L、（342.83±85.76）μg/L比（433.71±134.93）μg/L］，差异均有统计学意义（χ²=9.895，t=7.359、4.918、4.243；均P<0.05）。多因素logistic回归分析结果显示，伴有糖尿病及高HbA1c、PTH、SF水平是腹膜透析患者认知功能损害发生的独立危险因素（均P<0.05）。据此建立的列线图预测模型能有效预测腹膜透析患者认知功能损害的发生，校准曲线验证表明模型具有良好的预测准确性。结论　伴有糖尿病及高HbA1c、PTH、SF水平的腹膜透析患者认知功能损害发生的风险较高，应在治疗中加以重视并采取相应的预防措施。

关键词:

腹膜透析, 认知功能损害, 危险因素, logistic回归分析