Long-term and preoperative benzodiazepine use in older adults and risk for postoperative delirium

BACKGROUND Postoperative delirium (POD) is a common acute neurocognitive disorder characterised by sudden changes in mental status, including altered alertness, consciousness and cognition, and usually occurs in elderly patients. There is an urgent need to identify predictors of POD to prevent its onset, as it can significantly delay recovery from surgery. Benzodiazepines are among the most frequently prescribed medications, particularly for female individuals. They have a depressant effect on the central nervous system and are used to treat sleep disorders, anxiety, muscle relaxation and epilepsy. However, benzodiazepines may increase the risk of POD. OBJECTIVE We sought to investigate whether long-term benzodiazepine use, premedication and the interaction of benzodiazepines with sex are associated with the occurrence of POD and postoperative neurocognitive ability. DESIGN A part analysis of the prospective multicentre BioCog cohort study. SETTING Two-centre study conducted at Charité – Universitätsmedizin Berlin (Germany) and University Hospital Utrecht (Netherlands), both primary care hospitals, between October 2014 and September 2019. PATIENTS Data from 928 patients from the BioCog cohort study who underwent elective surgery were analysed. Of these, 42.3% were women, 18.6% reported long-term benzodiazepine use and 12.4% received premedication with benzodiazepines. MAIN OUTCOME MEASURES We studied the association of benzodiazepine use on cognition and the development of POD. We found that the timing of benzodiazepine use was crucial. RESULTS Long-term benzodiazepine use was significantly associated with the risk of developing POD, independent of sex ( P < 0.001). In contrast, premedication with benzodiazepines immediately before surgery was not associated with the risk of POD ( P = 0.242). males and females developed POD at similar rates. Regardless of sex, long-term benzodiazepine use elevated the risk of POD, unlike premedication with benzodiazepines. CONCLUSION Long-term use of benzodiazepines is associated with the development of POD, but short-term use as a premedicant is not. However, as this is an observational study, further research is needed to confirm these findings in a controlled setting. TRIAL REGISTRATION The study was registered at clinicaltrials.gov (NCT02265263).