Orthostatic hypotension status,influencing factors and related model construction in patients with Parkinson's disease

YANG Liu; LI Yachan; LIANG Xiao; BAI Xue; DU Chengxue; LIU Xiaomin; CAO Yuan; CHEN Ziming

Abstract：Objective To analyze the orthostatic hypotension (OH) status and its influencing factors in patients with Parkinson' s disease (PD) so as to construct an OH prediction model. Methods The clinical data of 208 PD patients admitted from January 2022 to March 2024 to the Affiliated Hospital of Northwest University were retrospectively analyzed. According to the occurrence of OH,they were divided into OH group and non-OH group. The baseline data such as age and discourse course were compared between the two groups. Logistic regression analysis was used to screen the independent influencing factors of OH in PD patients,and the prediction model of OH was constructed based on the results of the above regression analysis. Results The incidence of OH in 208 PD patients was 26.44％ (55/208). Age,disease course,Hoehn-Yahr grading,UPDRS Ⅲ score and incidence rates of RBD,vitamin D deficiency and circadian blood pressure rhythm revealed statistical differences between OH group and non-OH group (P<0.05). Logistic regression analysis showed that age[OR=12.368,95％CI (3.403-44.947),P<0.05],disease course[OR=14.517,95％CI (3.076-68.523),P<0.05]and reversed dipper circadian blood pressure rhythm[OR=3.150,95％CI (1.544-6.426),P<0.05]were independent risk factors for OH in PD patients. Based on the above three independent risk factors,a risk nomogram warning model was constructed. The internal validation of Bootstrap repeated sampling showed that the calibration curve was close to the ideal curve. Receiver operating characteristic (ROC) curve indicated that the area under the curve (AUC) was 0.880 (95％CI:0.828-0.931),and the model had good predictive efficiency. Conclusions PD patients with advanced age,5 years and above course of disease,and reversed dipper circadian blood pressure rhythm are more likely to develop OH. The prediction model constructed in this way can provide reference data for clinical management of PD.