Association of dyslipidemia and lipid-lowering treatment with risk of postoperative cognitive dysfunction (POCD): A systematic review and meta-analysis

Abstract

Introduction Postoperative cognitive dysfunction (POCD) is a common complication after surgery and involves a spectrum of symptoms spanning from mild to severe cognitive deficits that affect the daily lives of patients. Despite occurring in up to one third of older patients during the first few months after surgery, POCD is poorly characterized in terms of its epidemiology. Lipid imbalance in older age has frequently been linked to age-related cognitive impairment, but its role in determining patients’ risk of POCD is presently unknown. We therefore conducted a systematic review and meta-analysis of published research on the associations of dyslipidemia before surgery, use of lipid-lowering treatment before surgery and risk of POCD. Methods PubMed, Ovid SP and the Cochrane Database were searched for longitudinal studies that reported on associations of any measure of dyslipidemia (high total cholesterol, high low-density lipoprotein cholesterol, high triglycerides, low high-density lipoprotein; hypercholesterolemia) or on lipid-lowering treatment with POCD as relative risks (RR) or odds ratios. Any definition of POCD was permitted. Fixed-effects inverse variance models combined effects. Observational and intervention studies were considered separately. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) and Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines were followed and the search was registered on the PROSPERO database (CRD42017069013). Results Of 205 articles, 17 studies totaling 2725 patients (mean age 67 years) were included. All had observational design. Median follow-up period was 7 days (range 1 day to 4 years) and surgical procedures were most commonly of the heart or carotid artery. Studies focused almost exclusively on hypercholesterolemia (either not further defined, or defined through self-report or laboratory analysis) as a measure of dyslipidemia (n = 12 studies) and on statins as lipid-lowering treatment (n = 8 studies). Few other parameters of dyslipidemia were investigated. Of all included studies, only 2 had set out to address the present research question. All of the remaining investigations reported on pre-surgery exposure to dyslipidemia or lipid-lowering treatment and POCD risk in descriptive tables showing characteristics of patients with POCD compared with cognitively stable patients. Statistical adjustment for potential confounders was rare. In meta-analysis, we found no association of hypercholesterolemia with POCD risk (RR: 0.93; 95% CI: 0.80, 1.08; P = 0.34), whereas statin use before surgery was associated with a reduced POCD risk (RR: 0.81; 95% CI: 0.67, 0.98; P = 0.03). Information on duration of statin treatment was lacked throughout. Additional evidence from 2 trial studies that failed to meet inclusion criteria was conflicting. One reported better post-surgery cognitive function patients receiving rosuvastatin for 7 to 10 days before cardiac surgery compared with patients not receiving that treatment at 10 to 14-day follow-up but lacked a placebo–controlled group; the other of critically ill patients found no effect of statin treatment initiated immediately after surgery versus placebo on POCD risk at 6-month follow-up. Conclusion In this meta-analysis we found no association of hypercholesterolemia with risk of POCD, but we found that use of statins was associated with a reduced risk. Future epidemiological studies should consider collecting data on patients’ treatment history, including indication, drug type, dosage and treatment duration, and should strategically adjust for potential confounders in view to disentangle parameters of dyslipidemia and lipid-lowering treatment in POCD risk prediction. If those types of studies were to support the preliminary evidence presented here, trial studies could determine the causality in the reduced risk of POCD seen in statin users, as trials are inadequate at present in terms of study number and design.