Factors influencing resilience to postoperative delirium in adults undergoing elective orthopaedic surgery

Delirium occurs after elective arthroplasty in 17 per cent of adults1, and is associated with poor outcomes, including cognitive decline2, dementia3,4, and death5. Predisposing and precipitating risk factors accumulate and interact to precipitate delirium6. Much of the current literature analyses delirium as a dichotomous outcome, inevitably placing many people with symptoms of delirium, but falling short of a diagnosis, into the no-delirium group. Freedom from delirium symptoms should be investigated as an outcome. As evidence accumulates that delirium symptoms can also be associated with negative outcomes, it is important to identify the resilient groups in these studies and establish modifiable resilience predictors. Studies have explored risk factors for postoperative delirium; however, none to date has defined or considered delirium resilience as an outcome or phenotype. Resilience may be broadly defined as ‘the ability to withstand or recover quickly from difficult conditions’7,8. The aim of this study was to identify predictors of delirium resilience in the perioperative setting. As previously reported9,10, this observational cohort study recruited participants aged 65 years and over (without a diagnosis of dementia) due to undergo elective primary hip or knee replacement under spinal anaesthetic between March 2012 and October 2014. The study was performed in accordance with local ethics committee procedures, and all participants gave informed written consent (REC reference: 10/NIR01/5; protocol number: 09069PP-OPMS). Baseline demographic data, cognitive performance, and perioperative details were collected as previously described9,10. Patients were assessed for delirium once daily for the first three postoperative days using the Confusion Assessment Method (CAM)11, supported by the Mini Mental State Examination (MMSE)12, and nursing staff interviews. Postdischarge nursing and medical notes were interrogated where possible. Cerebrospinal fluid (CSF) and blood plasma samples were collected immediately preoperatively, as previously described9,10. Apolipoprotein E (APOE ε4) status and CSF biomarkers were analysed as described previously13 but were not analysed statistically in the context of this paper9,10. Two hundred and ninety-two participants with a preoperative MMSE score of 24 or more were included in this analysis, to prevent the inclusion of patients with undiagnosed dementia. Participants were categorized into ‘resilient’ or ‘non-resilient’ groups based on their postoperative MMSE and CAM scores. Delirium resilience was defined as a preoperative MMSE score of 24 or more, which did not subsequently decrease, maintaining or increasing original scores across all MMSE components, and not fulfilling any of the core CAM criteria, including acuity, inattention, altered level of consciousness, or disorganized thinking, during the first three postoperative days. An exception was made for the loss of one MMSE point in orientation, owing to the high frequency of ward movement during data collection. The preclinical covariates included in this analysis are summarized in Table 1. Logistic regression was carried out with resilience as the dependent variable. Variables were included based on statistical or clinical significance. The following independent variables were significant at the 5 per cent level in univariable analysis and included in the model: age; type of surgery; years in education; National Adult Reading Test (NART); Colour Trails 2; alcohol consumption; and CSF T-tau. Variables that were not statistically significant at this level but that were classed as clinically significant were also included: sex; Charlson Comorbidity Index; anticholinergic burden; Vertical Visual Analogue Pain Score (VVAS) for pain on movement; CSF Aβ1-42 concentration; and APOE ε4status. Several statistically or clinically significant variables were excluded owing to their correlation with other variables, or the low number of participants with available data. Analysis was performed using SPSS for Windows version 26 (IBM, Armonk, NY, USA). Methods and results are presented in accordance with STROBE guidance14, where possible. Baseline characteristics for the whole cohort, the resilient group and non-resilient group Values are n (%) unless otherwise indicated. Years in education assumed school starting age of 4 years. Alcohol units per week were estimated using the calculator at www.drinkaware.co.uk. Smoking status was recorded as current, ex-smoker, or non-smoker. Anticholinergic burden was calculated using the Ageing Brain Care tool at www.agingbraincare.org. *Student’s t test. †χ2 test. ‡Mann–Whitney U test. s.d., standard deviation; i.q.r., interquartile range; CCI, Charlson Comorbidity Index; GDS, Geriatric Depression Scale; VVAS, Vertical Visual Analgue Pain Score; NART, National Adult Reading Test; MMSE, Mini Mental State Examination; APOE ε4 , apolipoprotein E; CSF, cerebrospinal fluid; p-tau, phosphorylated tau; t-tau, total tau; Qalb, CSF to plasma albumin ratio; SBP, systolic blood pressure. Baseline characteristics are displayed in Table 1. Of the 292 participants included, 78 were categorized as resilient and 214 as non-resilient. The number of individuals included in the logistic regression analysis was less than the total number of study participants owing to missing data in certain variables. Of the 197 non-resilient individuals included in the logistic regression, 17 were delirious by CAM. The results of logistic regression analysis with resilience as the dependent variable are shown in Table 2. Age, NART score, VVAS pain on movement, and T-tau concentration were independent predictors of resilience to delirium in this cohort. The odds of being delirium-resilient reduced by 10 per cent (odds ratio (OR) 0.899) for each year increase in age, reduced by 2 per cent (OR 0.978) for each unit increase in VVAS score, and reduced by 0.4 per cent (OR 0.996) for each 10 ng/l increase in CSF t-tau concentration. Conversely, each unit increase in NART score increased the odds of resilience by 7 per cent (OR 1.065). Results of binary logistic regression analysis with independent predictors, using resilience as the predictor variable (n = 224) Model contains age at surgery, sex, hip or knee surgery, duration of education, Charlson Comorbidity Index (CCI), alcohol intake, National Audit Reading Test (NART) score, Vertical Visual Analogue Pain Score (VVAS) pain on movement, Preoperative Colour Trails 2 score, anticholinergic burden (ACB), Aβ1-42 concentration, T-tau concentration, and presence of APOE ε4. n/N , the number of people in the resilient category out of the total number of participants with data for this variable; OR, odds ratio; c.i., confidence interval; APOE ε4 , apolipoprotein E; T-tau, total tau; Ref. category, reference category. Younger age, higher NART score, lower preoperative pain score on movement, and lower concentration of CSF T-tau were independently associated with delirium resilience. Oldham et al. describe ‘pro-cognitive factors’ as baseline biopsychosocial factors that promote healthy cognitive function and predict delirium vulnerability15,16. Some participants were missing MMSE data in the current study, so a complete case analysis was conducted owing to concerns that multiple imputation may not be valid. The exclusion of some clinically important variables from the logistic regression model due to their correlation with other included variables reduced risk of skewing results but reduced the power of our analyses to detect true between-group differences. Devising the logistic regression model using both statistically and clinically significant variables may have also reduced the power of our analysis. The ceiling effect may provide limitation to our method of defining resilience. Those with high education levels or high preoperative MMSE score may experience undetected but meaningful cognitive decline. Higher late-life cognitive reserve is associated with reduced postoperative delirium incidence and severity17. Some people without delirium symptoms may have been placed into the non-resilient group as a result of using MMSE scores to define groups. Given the historical inclusion of people with delirium symptoms in control groups, we felt this was an appropriate risk. Further work will clarify consistent predictors of resilience. This work was funded by the Siew Keok Chin Scholarship, the Belfast Arthroplasty Research Trust (now TORCNI), and Belfast Trust Charitable Funds. E.L.C. has received grant funding from Alzheimer’s Research UK. D.F.M. has received grant funding from the NIHR RfPB programme for delirium research. E.M.L.B. is a PhD student at Queen’s University Belfast funded by the Department for the Economy (DfE). H.Z. is a Wallenberg Scholar supported by grants from the Swedish Research Council (#2018-02532); the European Research Council (#681712); Swedish State Support for Clinical Research (#ALFGBG-720931); the Alzheimer Drug Discovery Foundation (ADDF), USA (#201809-2016862); the AD Strategic Fund and the Alzheimer’s Association (#ADSF-21-831376-C, #ADSF-21-831381-C and #ADSF-21-831377-C); the Olav Thon Foundation; the Erling-Persson Family Foundation; Stiftelsen för Gamla Tjänarinnor, Hjärnfonden, Sweden (#FO2019-0228); the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 860197 (MIRIADE); and the UK Dementia Research Institute at UCL. E.M.L.B.: analysis and interpretation of data and drafting of manuscript; C.C.: analysis and interpretation of data, and drafting of manuscript; D.F.M.: conception and design of study, and revision of manuscript; B.M.: conception and design of study, and revision of manuscript; A.P.P.: conception and design of study, data acquisition, and revision of manuscript; D.B.: conception and design of study, data acquisition, and revision of manuscript; H.Z.: data acquisition and revision of manuscript; J.M.S.: data acquisition and revision of manuscript; E.L.C.: conception and design of the study, data acquisition, analysis and interpretation of data, revision of manuscript, and guarantor. The study was performed in accordance with local ethical committee procedures and all participants gave informed written consent (REC reference: 10/NIR01/5; protocol number: 09069PP-OPMS). HZ is a Wallenberg Scholar supported by grants from the Swedish Research Council (#2018-02532), the European Research Council (#681712), Swedish State Support for Clinical Research (#ALFGBG-720931), the Alzheimer Drug Discovery Foundation (ADDF), USA (#201809-2016862), the AD Strategic Fund and the Alzheimer’s Association (#ADSF-21-831376-C, #ADSF-21-831381-C and #ADSF-21-831377-C), the Olav Thon Foundation, the Erling-Persson Family Foundation, Stiftelsen för Gamla Tjänarinnor, Hjärnfonden, Sweden (#FO2019-0228), the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 860197 (MIRIADE), and the UK Dementia Research Institute at UCL. JMS acknowledges the support of the National Institute for Health Research University College London Hospitals Biomedical Research Centre, Wolfson Foundation, Alzheimer’s Research UK, Brain Research UK, Weston Brain Institute, Medical Research Council, British Heart Foundation, UK Dementia Research Institute and Alzheimer’s Association. Disclosure. HZ has served at scientific advisory boards for Eisai, Denali, Roche Diagnostics, Wave, Samumed, Siemens Healthineers, Pinteon Therapeutics, Nervgen, AZTherapies and CogRx, has given lectures in symposia sponsored by Cellectricon, Fujirebio, Alzecure and Biogen, and is a co-founder of Brain Biomarker Solutions in Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program (outside submitted work).

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.