ABSTRACT
Background: Postoperative delirium (POD) is a common complication after total hip replacement. This study aims to explore the relationship between preoperative fasting blood glucose (FBG) levels and POD in non-diabetic older patients undergoing total hip replacement. Materials and methods: This study included a total of 625 patients undergoing elective total hip replacement under combined spinal and epidural anesthesia from the PNDABLE study. The relationship between POD and preoperative FBG was analyzed by using the logistic regression model. The associations of FBG with individual cerebrospinal fluid (CSF) biomarkers were detected by using the multivariable linear regression model controlling for age, gender, and education level. The mediation effects were explored by mediation analyses with 5,000 bootstrap iterations, while sensitivity analysis was used to test the reliability and stability of the results. The receiver operating characteristic (ROC) curve and the nomogram model were applied to evaluate the efficacy of FBG and POD-related CSF biomarkers in predicting POD. POD assessment was performed two times daily by a trained anesthesiologist at 9:00-10:00 am and 2:00-3:00 pm on postoperative days 1-7 or before the patients were discharged from the hospital. POD was defined by the Confusion Assessment Method (CAM), and POD severity was measured using the Memorial Delirium Assessment Scale (MDAS). Enzyme-linked immunosorbent assay (ELISA) was used to measure CSF Aß40, Aß42, T-tau, and P-tau levels. Results: POD was detected in 10.2% (60/588) of the patients. Logistic regression analysis showed that after adjusting for age and education level, the increased levels of FBG (OR 1.427, 95% CI 1.117-1.824, P = 0.004), CSF P-tau (OR 1.039, 95% CI 1.019-1.060, P < 0.001), and CSF T-tau (OR 1.013, 95% CI 1.009-1.018, P < 0.001) were risk factors for POD, and the increased level of CSF Aß42 (OR 0.996, 95% CI 0.994-0.998, P = 0.001) was a protective factor for POD. Multivariable linear regression models showed that when adjusting for age, gender, and education level, in the POD group, higher preoperative FBG levels were negatively correlated with the CSF Aß42 level (ß = -0.290, P = 0.028) and positively correlated with CSF P-tau (ß = 0.384, P = 0.004) and T-tau (ß = 0.447, P < 0.001). In the non-POD group, a higher preoperative FBG was not related to CSF biomarkers. Mediated effect analysis showed that CSF T-tau (proportion = 17.1%) had an apparent mediation effect on the relationship between FBG and POD. Sensitivity analysis revealed that the results from the logistic regression and multivariable linear regression models were consistent with previous results. Conclusion: Increased preoperative FBG was a risk factor for POD in older patients without T2DM, and T-tau might mediate the relationship between FBG and POD.
