Preclinical Amyloid-β and Axonal Degeneration Pathology in Delirium

Abstract

Background: The clinical relevance of brain β-amyloidosis in older adults without dementia is not established. As delirium and dementia are strongly related, studies on patients with delirium may give pathophysiological clues.

Objective: To determine whether the Alzheimer’s disease (AD) cerebrospinal fluid (CSF) biomarkers amyloid-β 1-42 (Aβ₄₂), total tau (T-tau), and phosphorylated tau (P-tau) are associated with delirium in hip fracture patients with and without dementia.

Methods: CSF was collected in conjunction to spinal anesthesia in 129 patients. Delirium was assessed using the Confusion Assessment Method once daily in all patients, both pre- and postoperatively. The diagnosis of dementia at admission was based upon clinical consensus. CSF levels of Aβ₄₂, T-tau, and P-tau were analyzed.

Results: In patients without dementia, we found lower CSF Aβ₄₂ levels (median, 310 ng/L versus 489 ng/L, p = 0.006), higher T-tau levels (median, 505 ng/L versus 351 ng/L, p = 0.02), but no change in P-tau in patients who developed delirium (n = 16) compared to those who remained lucid (n = 49). Delirious patients also had lower ratios of Aβ₄₂ to T-tau (p < 0.001) and P-tau (p = 0.001) relative to those without delirium. CSF Aβ₄₂ and T-tau remained significantly associated with delirium status in adjusted analyses. In patients with dementia, CSF biomarker levels did not differ between those with (n = 54) and without delirium (n = 10).

Conclusion: The reduction in CSF Aβ₄₂, indicating β-amyloidosis, and increase in T-tau, indicating neurodegeneration, in hip fracture patients without dementia developing delirium indicates that preclinical AD brain pathology is clinically relevant and possibly plays a role in delirium pathophysiology.

Keywords

Alzheimer’s disease biomarkers cerebrospinal fluid delirium dementia physiopathology

INTRODUCTION

More than 30% of individuals over 65 years of age have biomarker evidence of brain β-amyloidosis without having dementia [1], but it is not settled whether this represents preclinical Alzheimer’s disease (AD) or just is a bystander of aging. Delirium, an acute and fluctuating reduction in attention, awareness, and cognition, may arise in older people without dementia when exposed to external stressors, e.g., trauma and surgery. Clinical studies on older hip fracture patients developing delirium may therefore give pathophysiological clues on the relevance and clinical consequences of preclinical brain β-amyloidosis.

Cognitive impairment and dementia are strong risk factors for delirium [2, 3], and delirium is independently associated with an increased incidence of chronic cognitive decline and dementia [2, 4]. Thus, delirium and dementia are closely interrelated, and common pathophysiological mechanisms are believed to exist [2, 5]. The pathophysiological links are, however, not known [2].

Amyloid-β 1-42 (Aβ₄₂), total tau (T-tau), and phosphorylated tau (P-tau) are the core AD cerebrospinal fluid (CSF) biomarkers, and are believed to reflect brain pathophysiology. Reduced CSF Aβ₄₂ concentration reflects accumulation of aggregated amyloid-β (Aβ) in amyloid plaques [6, 7]. Increased T-tau level reflects the intensity of axonal degeneration [8, 9], while increased P-tau level correlates with the amount of tangle pathology in the brain [10]. These CSF biomarkers may predict the occurrence of delirium, but studies so far show conflicting results [11, 12]. However, these two studies excluded patients with a dementia diagnosis so it is difficult to draw conclusions. The aim of our study was to examine if the AD CSF biomarkers are associated with delirium in hip fracture patients with and without dementia.

MATERIALS AND METHODS

Participants

Patients were recruited from the Oslo Orthogeriatric Trial (OOT): a randomized, controlled trial comparing orthogeriatric care with usual orthopedic care. Delirium incidence was a secondary outcome, and there was no difference in delirium rates between intervention and control group. A predefined secondary aim was to collect CSF in conjunction with spinal anesthesia in order to elucidate pathophysiological mechanisms in delirium. The OOT assessed for eligibility all patients that were acutely admitted to the Ullevaal Clinic of Oslo University Hospital with a hip fracture from September 2009 to January 2012. Patients were excluded if regarded as moribund and if the hip fracture was part of a high-energy trauma. Details have been published previously [13, 14].

Assessments

Relatives or health professionals were interviewed regarding pre-fracture status. Pre-fracture cognitive status was assessed using the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE) [15, 16], and pre-fracture Activities of Daily Living (ADL) performance was assessed by the Barthel Index [17]. Acute Physiology and Chronic Health Evaluation (APACHE) II score [18] at admittance and the American Society of Anesthesiologists (ASA) score [19] before hip fracture surgery were also recorded. Delirium was assessed once daily by the study geriatrician (LOW) or a study nurse using the Confusion Assessment Method (CAM) [20], both preoperatively and postoperatively. All patients were screened until the fifth postoperative day, and patients with delirium were screened until discharge. CAM scores were based on an interview with the patient, including tests of cognition, attention, and alertness (digit span test, orientation, and delayed recall), information from close relatives and nurses, and review of hospital records from the previous 24 hours. Delirium assessment was done regularly only on weekdays, but staff members who had been working during the weekends were interviewed every Monday, and case notes were reviewed in order to ascertain potential episodes of delirium. The diagnosis of dementia at admission was based upon consensus between an experienced old age psychiatrist (KE) and an experienced geriatrician (TBW) who independently assessed whether the hip fracture patients fulfilled the ICD-10 criteria [21] for dementia. They were allowed access to relevant information extracted from clinical records, e.g., previous dementia diagnoses and cognitive test results, as well as scores on IQCODE, Clinical Dementia Rating (CDR) [22], and Nottingham Extended ADL Scale (NEADL) [23] at admission and cognitive test results, CDR score, and NEADL score at one-year follow-up. They were blind to records of delirium status during the hospital stay. Inter-rater agreement was very good (kappa 0.87), and disagreements were resolved through discussion.

Cerebrospinal fluid collection and analyses

CSF samples were taken by the anesthesiologist immediately before injection of the anesthetic agent. CSF was collected in polypropylene tubes, centrifuged, the supernatant aliquoted in polypropylene vials, frozen as soon as possible, and stored at –80°C. CSF was thawed, aliquoted once more in polypropylene vials, and frozen again before it was sent for analyses.

In December 2013, the frozen supernatant samples of CSF were sent on dry ice to the Clinical Neurochemistry Laboratory at Sahlgrenska University Hospital, Mölndal, Sweden, for analyses. Aβ₄₂, T-tau, and P-tau concentrations were determined using INNOTEST enzyme-linked immunosorbent assays (Fujirebio, Ghent, Belgium). The INNOTEST Aβ₄₂ method employs antibodies that specifically detect the neo-epitopes of the first and last amino acid of the 42 amino acid-long Aβ sequence [24]. All analyses were performed by board-certified laboratory technicians, who were masked to clinical data, using one batch of reagents with intra-assay coefficients of variation below 10%. Values outside the quantification limits were imputed as the quantification limit, 125 ng/L for Aβ₄₂ (4 patients) and 1380 ng/L for T-tau (3 patients).

Statistical analyses

This was an exploratory study in a patient cohort from a randomized controlled trial, and thus no power analysis was done for this specific study. Due to a non-normal distribution of the data, Mann-Whitney U-tests were used for analyses of continuous variables. Chi-Square or Fisher’s Exact Tests were used for analyses of categorical variables. Significance was set at p < 0.05. We calculated ratios of Aβ₄₂ to T-tau and P-tau because they may increase the predictive value of the CSF biomarkers [25]. Subjects were classified as Aβ₄₂+ (<530 ng/L) or Aβ₄₂- (≥530 ng/L), T-tau+ (>350 ng/L) or T-tau- (≤350 ng/L), and P-tau+ (≥60 ng/L) or P-tau- (<60 ng/L) according to cut-off values suggested by Hansson et al. [26]. In the dementia-free stratum, logistic regression analyses were conducted to adjust for potential confounding of the association between biomarker concentrations and delirium. All statistical analyses were performed using SPSS Statistics version 21 (IBM, Armonk NY).

Ethical considerations

The study was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from the patients. If a patient was unable to give an informed consent, a presumed consent in combination with written assent from the closest relative was obtained. The study was approved by the Regional Committee for Ethics in Medical Research in Norway (REK S-09169a).

RESULTS

CSF samples were obtained from 143 of 332 patients included in OOT. 14 patients were later excluded (Supplementary Figure 1). Demographics, delirium incidence, IQCODE scores, and proportion with dementia were not statistically significantly different in the 129 patients left for present analyses compared with the remaining OOT patients.

Dementia versus no dementia in the full sample

Demographics and CSF biomarker characteristics of the full sample are given in Supplementary Table 1. 54 out of 64 (84%) patients with dementia developed delirium in conjunction to the hip fracture, either preoperatively or postoperatively, compared to 16 out of 65 (25%) among patients without dementia (p < 0.001). Abnormal CSF biomarker levels were more prevalent in patients with dementia than in patients without dementia: 59/64 (92%) versus 41/65 (63%) were Aβ₄₂+, and 43/64 (67%) versus 36/65 (55%) were T-tau+. Thus, further analyses were stratified on dementia status.

Delirium versus no delirium in patients without dementia

Demographics for patients without dementia are given in Table 1. Patients who developed delirium were comparable to those who did not develop delirium with respect to age, gender, IQCODE score, prevalence of nursing home residency, ADL score, and morbidity. There was a trend toward higher APACHE score in patients with delirium compared to those without. CSF Aβ₄₂ levels were 179 ng/L (37%) lower in patients who developed delirium compared to those who did not, and 81% of patients with delirium were Aβ₄₂+ (Table 2 and Fig. 1A). CSF T-tau levels were 154 ng/L (44%) higher in patients with delirium, but the difference in proportion of patients being T-tau+ did not reach statistical significance (Table 2 and Fig. 1B). CSF P-tau levels were 18 ng/L (33%) higher in patients with delirium, however the difference was only borderline significant. There was a trend toward a higher proportion of P-tau+ patients in patients who developed delirium (Table 2). Delirious patients also had lower ratios of Aβ₄₂ to T-tau and P-tau relative to those without delirium (Table 2 and Fig. 1C). When excluding 9 patients with preoperative delirium from the analyses, the differences in biomarker levels became slightly more pronounced (Supplementary Table 2); 251 ng/L for Aβ₄₂, 185 ng/L for T-tau, and 21 ng/L for P-tau. CSF T-tau and P-tau levels were significantly higher (Supplementary Table 2), and the ratios of Aβ₄₂ to T-tau and P-tau significantly lower, in patients who developed delirium postoperatively compared to those who did not (data not shown). The difference in CSF Aβ₄₂ levels did not reach statistical significance in this subgroup analysis (p = 0.09). When excluding patients ≤75 years old, differences in biomarker levels showed a similar pattern (datanot shown).

Delirium versus no delirium in patients with dementia

Demographics for patients with dementia are given in Table 1. No CSF biomarker showed statistically significant differences between patients with and without delirium in the dementia stratum (Table 2). The results were the same after excluding patients with preoperative delirium (data not shown).

Preoperative delirium versus postoperative delirium

Delirium continued postoperatively in all patients with preoperative delirium. In the non-demented stratum, nine patients developed delirium preoperatively, while seven patients developed delirium postoperatively. The CSF biomarker levels were not significantly different between patients who developed delirium before or after surgery (data not shown). This result was the same in the full sample (43/65 [66%] developed delirium preoperatively) and in the demented stratum (34/49 [69%] developed delirium preoperatively) (data not shown). Preoperative delirium status was missing in five patients with dementia.

Multivariable analyses

Because the degree of cognitive impairment might have confounded the association between biomarkers and delirium despite the stratification procedure, we built separate regression models for each of the CSF biomarkers and biomarker ratios entering the biomarker/ratio, age, gender, and IQCODE score into logistic regression models with delirium status as the dependent variable. These analyses were limited to the dementia-free stratum, where the associations had been observed in bivariate analyses. CSF levels of Aβ₄₂ and T-tau, and the ratios of Aβ₄₂ to Tau and P-tau, remained significantly associated with delirium status in these analyses. CSF P-tau levels were not significantly associated with delirium in the adjusted model (Table 3).

Combinations of amyloid and tau pathology

Patients were further divided into four groups according to the presence of amyloid pathology (Aβ₄₂+ or Aβ₄₂–) and/or tau pathology (P-tau+ orP-tau–). In the stratum without dementia, 10/18 (56%) of the patients with evidence of both amyloid and tau pathology developed delirium, while the three other groups had a delirium occurrence below 20% (Supplementary Figure 2A). In the stratum with dementia, around 85% of the Aβ₄₂+ patients developed delirium (Supplementary Figure 2B). A similar figure for the full sample is presented in Supplementary Figure 2C.

DISCUSSION

This is, to our knowledge, the first study to examine the relationship between preoperative CSF levels of Aβ₄₂, T-tau, and P-tau and delirium in both patients with and without dementia. It is also the first study to show a convincing association between these biomarkers and delirium. In patients without dementia, we found lower CSF Aβ₄₂ levels and also lower ratios of Aβ₄₂ to T-tau and P-tau, together with an increase in CSF T-tau levels in hip fracture patients who developed delirium compared to those who remained delirium free.

Strengths of our study include the daily and bedside assessment of delirium and the inclusion of patients both with and without dementia. The ascertainment of delirium and dementia was based on validated instruments. A larger number of patients included than in other delirium biomarker studies made subgroup analyses on dementia status and pre- and postoperative delirium possible. CSF was analyzed in a laboratory with an extensive experience on Aβ₄₂, T-tau, and P-tau analyses in CSF.

Some limitations deserve a comment. In the stratum without dementia, we may not have the power to show a difference even if one is there since our sample size is small and the differences are also smaller due to the generally more pathological biomarker levels. Dementia diagnoses based on a cognitive assessment before the patients became hospitalized would be better than our retrospective dementia diagnoses. However, since hip fracture patients are acutely admitted to the hospital, this was not feasible. We also believe that our consensus diagnosis of dementia is more accurate than merely IQCODE cut-offs. The dementia diagnosis is, however, not etiological, and thus even though a high proportion of the patients with dementia likely have AD dementia, they may also have other dementias.

The biomarkers and biomarker ratios remained significantly associated with delirium in hip fracture patients without dementia after adjusting for age, gender, and IQCODE score. This indicates that preclinical AD pathology increases the risk of delirium before the patient experiences a measurable chronic cognitive impairment. It is well established that reduced CSF Aβ₄₂ concentration reflects accumulation of aggregated Aβ in amyloid plaques in the brain [6, 7], also in cognitively normal individuals [27, 28], but there is very limited data on whether amyloid pathology in older people without dementia has any clinical consequences. An important finding is therefore that older patients without dementia with clinically silent amyloid pathology were found to develop symptoms (i.e., delirium) when exposed to physical stress. Amyloid pathology has been associated with dysfunction in brain networks in cognitively normal individuals [29 –33], and a breakdown of network connectivity in the brain has been suggested to be the neurobiological substrate of delirium [34]. It is likely that axonal degeneration and tangle pathology reflected by CSF T-tau and P-tau also have detrimental effects on brain network connectivity [33]. We found that patients without dementia with positive Aβ₄₂ and P-tau were more likely to develop delirium than patients with an abnormality in only one of these two biomarkers; thus, the vulnerability to delirium seems to increase with an increasing degree of neuropathology. This is in accordance with the results of Davis and colleagues, which showed that the risk of delirium-like behavior increased with increasing brain pathology in mice [35].

Although delirium is an independent risk factor for incident dementia [2, 4], our findings raise the question of whether delirium is simply an early symptom of a preclinical dementia. However, Cavallari and colleagues did not find an association between magnetic resonance imaging (MRI) measures of white-matter damage, global brain, and hippocampal volume and the incidence of delirium in a population of patients without dementia [36]. A possible explanation could be that most patients without dementia have insufficient structural brain pathology at a macroscopic level to be detected by MRI. CSF biomarkers may have a greater potential for detecting early brain pathology relevant for delirium.

Acute events may also change CSF biomarker levels. The literature on acute biomarker changes is sparse. However, CSF Aβ levels have been shown to decrease within hours after intake of 60 mg citalopram [37], CSF Aβ₄₂ levels are decreased in acute purulent bacterial meningitis and increase after antibiotic treatment [38], and CSF T-tau, but not P-tau, levels increase transiently after acute stroke [8]. It is therefore possible that trauma and stress in relation to the hip fracture could have altered the biomarker levels, but this would not explain different CSF biomarker levels in patients with and without delirium, since all patients experienced a hip fracture. The biomarker changes may also be related to pathophysiological processes in delirium per se, e.g., acute neurodegeneration could be a pathophysiological process in delirium resulting in increased CSF T-tau levels.

We found no significant associations between the CSF biomarkers and delirium in hip fracture patients with dementia. Patients with dementia have more extensive brain pathology, and in our study almost all patients with dementia had abnormal CSF biomarkers and developed delirium. The abnormalities in biomarkers have likely reached a plateau in dementia [39], thus the differences in biomarker levels between patients with dementia are likely to be small.

Given that abnormal CSF biomarkers constitute a risk factor for delirium, one might hypothesize that patients with such abnormalities require a less noxious insult before developing delirium. Thus, we examined if patients who developed delirium preoperatively had more abnormal CSF levels than those who did not develop the syndrome until after surgery. We did not find any such differences. If delirium as such causes acute biomarker changes, changes could occur at different time points of the delirium course, for example before delirium develops or as a result of delirium. Thus, the lack of difference in biomarker levels between patients who developed delirium preoperatively and postoperatively in our study suggests one of two scenarios: either that no acute biomarker changes occur before or during delirium, or that acute changes occur before the patient develops delirium and stay altered during delirium.

Only two previous studies have examined the association between the AD CSF biomarkers and delirium. Witlox and colleagues performed a prospective cohort study in older hip fracture patients similar to our study [11]. However, they excluded patients with a dementia diagnosis and patients not capable of providing informed consent, and no patients in their cohort developed delirium preoperatively. They found no association between biomarkers and delirium. Biomarker levels in their study, both in patients with and without delirium, were much closer to the normal values than we found (Supplementary Table 2). This might indicate that AD brain pathology was less prevalent in their study population. Xie and colleagues found CSF levels of Aβ₄₂ and T-tau more similar to our findings [12]. Like Witlox and colleagues, they did not find an association between preoperative CSF biomarker levels and postoperative delirium in a population ≥63 years old having elective total hip or knee replacement. However, when they divided the participants into quartiles according to Aβ₄₀/T-tau and Aβ₄₂/Tau ratios, they found a higher delirium incidence in the lowest quartile.

Conclusions

We demonstrate a reduction in CSF Aβ₄₂, indicating β-amyloidosis, together with an increase in T-tau, indicating neurodegeneration, in hip fracture patients without dementia developing delirium. These findings indicate that preclinical brain amyloidosis has clinical consequences and suggest that these patients have preclinical AD. Our findings also suggest that AD brain pathology plays a role in delirium pathophysiology. Larger studies of different populations are needed to more firmly assess the potential association between Aβ₄₂, T-tau and P-tau and delirium.

Footnotes

ACKNOWLEDGMENTS

The authors would like to thank Knut Engedal, MD, PhD (University of Oslo and Vestfold Hospital Trust) for his contribution to the dementia consensus diagnoses, and Elisabeth Fragaat (Oslo University Hospital) and Tone Fredriksen (Oslo University Hospital) for collection of data. The authors would also like to thank the study participants, and acknowledge the contributions of the Department of Orthopedic Surgery and the Department of Anesthesiology at Oslo University Hospital, Norway.

The study was mainly funded by the Research Council of Norway through the program “Improving mental health of older people through multidisciplinary efforts” (grant no. 187980/H10). Further, we have received funding from Oslo University Hospital, The Sophies Minde Foundation, The Norwegian Association for Public Health, Civitan’s Research Foundation, South-Eastern Norway Regional Health Authority, the Medical Student Research Program at the University of Oslo, the Knut and Alice Wallenberg Foundation, the Swedish Research Council, and the Torsten Söderberg Foundation at the Royal Swedish Academy of Sciences.

The funders had no role in the design, methods, subject recruitment, data collections, analysis, or preparation of paper.

Authors’ disclosures available online ().

Appendix

References

Jack

, Wiste

, Weigand

, Rocca

, Knopman

, Mielke

, Lowe

, Senjem

, Gunter

, Preboske

, Pankratz

, Vemuri

, Petersen

(2014) Age-specific population frequencies of cerebral beta-amyloidosis and neurodegeneration among people with normal cognitive function aged 50-89 years: A cross-sectional study. Lancet Neurol 13, 997–1005.

Fong

, Davis

, Growdon

, Albuquerque

, Inouye

(2015) The interface between delirium and dementia in elderly adults. Lancet Neurol 14, 823–832.

Ahmed

, Leurent

, Sampson

(2014) Risk factors for incident delirium among older people in acute hospital medical units: A systematic review and meta-analysis. Age Ageing 43, 326–333.

Davis

, Muniz Terrera

, Keage

, Rahkonen

, Oinas

, Matthews

, Cunningham

, Polvikoski

, Sulkava

, MacLullich

, Brayne

(2012) Delirium is a strong risk factor for dementia in the oldest-old: A population-based cohort study. Brain 135, 2809–2816.

Eikelenboom

, Hoogendijk

WJG

(1999) Do delirium and Alzheimer’s dementia share specific pathogenetic mechanisms? Dement Geriatr Cogn Disord 10, 319–324.

Strozyk

, Blennow

, White

, Launer

(2003) CSF Abeta 42 levels correlate with amyloid-neuropathology in a population-based autopsy study. Neurology 60, 652–656.

Blennow

, Mattsson

, Scholl

, Hansson

, Zetterberg

(2015) Amyloid biomarkers in Alzheimer’s disease. Trends Pharmacol Sci 36, 297–309.

Hesse

, Rosengren

, Andreasen

, Davidsson

, Vanderstichele

, Vanmechelen

, Blennow

(2001) Transient increase in total tau but not phospho-tau in human cerebrospinal fluid after acute stroke. Neurosci Lett 297, 187–190.

Riemenschneider

, Wagenpfeil

, Vanderstichele

, Otto

, Wiltfang

, Kretzschmar

, Vanmechelen

, Forstl

, Kurz

(2003) Phospho-tau/total tau ratio in cerebrospinal fluid discriminates Creutzfeldt-Jakob disease from other dementias. Mol Psychiatry 8, 343–347.

10.

Tapiola

, Alafuzoff

, Herukka

, Parkkinen

, Hartikainen

, Soininen

, Pirttila

(2009) Cerebrospinal fluid beta-amyloid 42 and tau proteins as biomarkers of Alzheimer-type pathologic changes in the brain. Arch Neurol 66, 382–389.

11.

Witlox

, Kalisvaart

, de Jonghe

, Verwey

, van Stijn

, Houdijk

, Traast

, Maclullich

, van Gool

, Eikelenboom

(2011) Cerebrospinal fluid beta-amyloid and tau are not associated with risk of delirium: A prospective cohort study in older adults with hip fracture. J Am Geriatr Soc 59, 1260–1267.

12.

Xie

, Swain

, Ward

SAP

, Zheng

, Dong

, Sunder

, Burke

, Escobar

, Zhang

, Marcantonio

(2014) Preoperative cerebrospinal fluid β-Amyloid/Tau ratio and postoperative delirium. Ann Clin Transl Neurol 1, 319–328.

13.

Wyller

, Watne

, Torbergsen

, Engedal

, Frihagen

, Juliebo

, Saltvedt

, Skovlund

, Raeder

, Conroy

(2012) The effect of a pre- and post-operative orthogeriatric service on cognitive function in patients with hip fracture. The protocol of the Oslo Orthogeriatrics Trial. BMC Geriatr 12, 36.

14.

Watne

, Torbergsen

, Conroy

, Engedal

, Frihagen

, Hjorthaug

, Juliebo

, Raeder

, Saltvedt

, Skovlund

, Wyller

(2014) The effect of a pre- and postoperative orthogeriatric service on cognitive function in patients with hip fracture: Randomized controlled trial (Oslo Orthogeriatric Trial). BMC Med 12, 12.

15.

Jorm

(1994) A short form of the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE): Development and cross-validation. Psychol Med 24, 145–153.

16.

Jackson

, MacLullich

, Gladman

, Lord

, Sheehan

(2016) Diagnostic test accuracy of informant-based tools to diagnose dementia in older hospital patients with delirium: A prospective cohort study. Age Ageing 45, 505–511.

17.

Mahoney

, Barthel

(1965) Functional evaluation: The Barthel Index. Md State Med J 14, 61–65.

18.

Knaus

, Draper

, Wagner

, Zimmerman

(1985) APACHE II: A severity of disease classification system. Crit Care Med 13, 818–829.

19.

American Society of Anesthesiologists (1963) New classification of physical status. Anesthesiology 24, 111.

20.

Inouye

, van Dyck

, Alessi

, Balkin

, Siegal

, Horwitz

(1990) Clarifying confusion: The confusion assessment method. A new method for detection of delirium. Ann Intern Med 113, 941–948.

21.

World Health Organization (1992) The ICD-10 classification of mental and behavioural disorders: Clinical descriptions and diagnostic guidelines. World Health Organization, Geneva.

22.

Hughes

, Berg

, Danziger

, Coben

, Martin

(1982) A new clinical scale for the staging of dementia. Br J Psychiatry 140, 566–572.

23.

Lincoln

, Gladman

(1992) The Extended Activities of Daily Living scale: A further validation. Disabil Rehabil 14, 41–43.

24.

Andreasen

, Hesse

, Davidsson

, Minthon

, Wallin

, Winblad

, Vanderstichele

, Vanmechelen

, Blennow

(1999) Cerebrospinal fluid beta-amyloid(1-42) in Alzheimer disease: Differences between early- and late-onset Alzheimer disease and stability during the course of disease. Arch Neurol 56, 673–680.

25.

Holtzman

(2011) CSF biomarkers for Alzheimer’s disease: Current utility and potential future use. Neurobiol Aging 32, S4–S9.

26.

Hansson

, Zetterberg

, Buchhave

, Londos

, Blennow

, Minthon

(2006) Association between CSF biomarkers and incipient Alzheimer’s disease in patients with mild cognitive impairment: A follow-up study. Lancet Neurol 5, 228–234.

27.

Landau

, Lu

, Joshi

, Pontecorvo

, Mintun

, Trojanowski

, Shaw

, Jagust

(2013) Comparing positron emission tomography imaging and cerebrospinal fluid measurements of beta-amyloid. Ann Neurol 74, 826–836.

28.

Fagan

, Mintun

, Shah

, Aldea

, Roe

, Mach

, Marcus

, Morris

, Holtzman

(2009) Cerebrospinal fluid tau and ptau(181) increase with cortical amyloid deposition in cognitively normal individuals: Implications for future clinical trials of Alzheimer’s disease. EMBO Mol Med 1, 371–380.

29.

Sheline

, Raichle

, Snyder

, Morris

, Head

, Wang

, Mintun

(2010) Amyloid plaques disrupt resting state default mode network connectivity in cognitively normal elderly. Biol Psychiatry 67, 584–587.

30.

Drzezga

, Becker

, Van Dijk

, Sreenivasan

, Talukdar

, Sullivan

, Schultz

, Sepulcre

, Putcha

, Greve

, Johnson

, Sperling

(2011) Neuronal dysfunction and disconnection of cortical hubs in non-demented subjects with elevated amyloid burden. Brain 134, 1635–1646.

31.

Hedden

, Van Dijk

, Becker

, Mehta

, Sperling

, Johnson

, Buckner

(2009) Disruption of functional connectivity in clinically normal older adults harboring amyloid burden. J Neurosci 29, 12686–12694.

32.

Mormino

, Smiljic

, Hayenga

, Onami

, Greicius

, Rabinovici

, Janabi

, Baker

, Yen

, Madison

, Miller

, Jagust

(2011) Relationships between beta-amyloid and functional connectivity in different components of the default mode network in aging. Cereb Cortex 21, 2399–2407.

33.

Wang

, Brier

, Snyder

, Thomas

, Fagan

, Xiong

, Benzinger

, Holtzman

, Morris

, Ances

(2013) Cerebrospinal fluid Aβ42, phosphorylated Tau181, and resting-state functional connectivity. JAMA Neurol 70, 1242–1248.

34.

Sanders

(2011) Hypothesis for the pathophysiology of delirium: Role of baseline brain network connectivity and changes in inhibitory tone. Med Hypotheses 77, 140–143.

35.

Davis

, Skelly

, Murray

, Hennessy

, Bowen

, Norton

, Brayne

, Rahkonen

, Sulkava

, Sanderson

, Rawlins

, Bannerman

, MacLullich

, Cunningham

(2015) Worsening cognitive impairment and neurodegenerative pathology progressively increase risk for delirium. Am J Geriatr Psychiatry 23, 403–415.

36.

Cavallari

, Hshieh

, Guttmann

, Ngo

, Meier

, Schmitt

, Marcantonio

, Jones

, Kosar

, Fong

, Press

, Inouye

, Alsop

(2015) Brain atrophy and white-matter hyperintensities are not significantly associated with incidence and severity of postoperative delirium in older persons without dementia. Neurobiol Aging 36, 2122–2129.

37.

Sheline

, West

, Yarasheski

, Swarm

, Jasielec

, Fisher

, Ficker

, Yan

, Xiong

, Frederiksen

, Grzelak

, Chott

, Bateman

, Morris

, Mintun

, Lee

, Cirrito

(2014) An antidepressant decreases CSF Aβ production in healthy individuals and in transgenic AD mice. Sci Transl Med 6, 8–.

38.

Sjogren

, Gisslen

, Vanmechelen

, Blennow

(2001) Low cerebrospinal fluid beta-amyloid 42 in patients with acute bacterial meningitis and normalization after treatment. Neurosci Lett 314, 33–36.

39.

Jack

, Knopman

, Jagust

, Petersen

, Weiner

, Aisen

, Shaw

, Vemuri

, Wiste

, Weigand

, Lesnick

, Pankratz

, Donohue

, Trojanowski

(2013) Tracking pathophysiological processes in Alzheimer’s disease: An updated hypothetical model of dynamic biomarkers. Lancet Neurol 12, 207–216.