Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Jun 1.
Published in final edited form as: Neurobiol Aging. 2015 Feb 28;36(6):2122–2129. doi: 10.1016/j.neurobiolaging.2015.02.024

Brain atrophy and white matter hyperintensities are not significantly associated with incidence and severity of postoperative delirium in older persons without dementia

Michele Cavallari a,1, Tammy T Hshieh b,c,1, Charles RG Guttmann a, Long H Ngo d, Dominik S Meier a, Eva M Schmitt c, Edward R Marcantonio d, Richard N Jones e,f, Cyrus M Kosar c, Tamara G Fong c,g, Daniel Press g, Sharon K Inouye c,d,2, David C Alsop, on behalf of the SAGES Study Grouph,2,*
PMCID: PMC4433616  NIHMSID: NIHMS671513  PMID: 25824618

Abstract

Postoperative delirium is a common complication in older people, and is associated with increased mortality, morbidity, institutionalization and caregiver burden. Although delirium is an acute confusional state characterized by global impairments in attention and cognition, it has been implicated in permanent cognitive impairment and dementia. The pathogenesis of delirium, as well as the mechanisms leading to these disabling consequences, remains unclear.

The present study is the first to address the potential predisposing role of brain morphological changes towards postoperative delirium in a large prospective cohort of patients undergoing elective surgery using state-of-the-art magnetic resonance imaging (MRI) techniques conducted before admission. We investigated the association of MRI-derived quantitative measures of white matter damage, global brain and hippocampal volume with the incidence and severity of delirium.

Pre-surgical white matter hyperintensities (WMH), whole brain and hippocampal volume were measured in 146 consecutively enrolled subjects, ≥70 years old, without dementia who were undergoing elective surgery. These three pre-surgical MRI indices were tested as predictors of incidence and severity of subsequent delirium.

Out of 146 subjects, 32 (22%) developed delirium. We found no statistically significant differences in WMH, whole brain or hippocampal volume between subjects with and without delirium. Both unadjusted and adjusted (age, gender, vascular comorbidity, general cognitive performance) regression analyses demonstrated no statistically significant association between any of the MRI measures with respect to delirium incidence or severity.

In persons without dementia, preexisting cerebral white matter hyperintensities, general and hippocampal atrophy may not predispose to postoperative delirium or worsen its severity.

Keywords: Delirium, brain atrophy, hippocampal atrophy, white matter hyperintensities, neuroimaging

1. Introduction

Delirium is an acute confusional state marked by global impairments in attention and cognition (Inouye et al., 2014a). Common yet under-diagnosed, postoperative delirium occurs in 11–51% of hospitalized older patients (Inouye et al., 2014a), and significantly increases their risk for death, institutionalization, and dementia (Witlox et al., 2010). Mortality rates among hospitalized patients who develop delirium are as high as those among patients with myocardial infarction or sepsis (Inouye, 2006).

Despite its clinical impact, the pathophysiology of delirium remains poorly understood. It is believed that brain pathology plays an important role. Preexisting cognitive impairment and dementia are important predisposing factors for delirium. In addition, many precipitating factors leading to delirium, such as electrolyte disturbances, malnutrition, infections, anemia, shock, and psychotropic medications have been identified (Inouye, 2006; Inouye et al., 2014a). Many cases of delirium are preventable through management of these precipitating factors.

A pathophysiological role for preexisting brain damage with respect to delirium incidence has been hypothesized. The hypothesis is based upon evidence that primary neurologic diseases (e.g., dementia, stroke, traumatic brain injury) are risk factors for delirium (Inouye, 2006). Neuroimaging studies showing structural and functional magnetic resonance imaging (MRI) abnormalities in subjects with delirium further support this hypothesis (Alsop et al., 2006; Hughes et al., 2012; Lipowski, 1990; Soiza et al., 2008; Trzepacz, 2000). As long as three decades ago, brain atrophy was found to be associated with delirium in computed tomography studies. Koponen et al. found that delirious patients had more cortical atrophy, wider Sylvian fissures and more prominent ventricular dilatation on computed tomography (Koponen et al., 1989, 1987). Global and regional perfusion abnormalities have been also observed on brain imaging of delirious patients (Fong et al., 2006; Siepe et al., 2011). As neuroimaging advanced, technology has become available to characterize the chronology of structural and functional brain abnormalities related to delirium, and provided clues to the pathogenesis of delirium and associated cognitive decline (Hughes et al., 2012; Saczynski et al., 2012). MRI-derived measures of global and regional brain atrophy, as well as cerebral white matter damage, have been associated with occurrence of delirium (for a systematic review see Alsop et al., 2006). Two small MRI studies found higher cerebral white matter hyperintensity (WMH) burden in subjects who developed delirium after electroconvulsive therapy (Figiel et al., 1990a, 1990b). A few studies suggested a predictive role for multiple brain infarcts (Otomo et al., 2013) and cerebral white matter damage, as measured by WMH burden or diffusion tensor imaging (DTI), for delirium incidence after cardiac surgery (Hatano et al., 2013; Root et al., 2013; Shioiri et al., 2010). In addition, recent resting-state functional MRI findings during and after an episode of delirium suggested that changes in cortical and subcortical functional connectivity underlie the pathogenesis of delirium (Choi et al., 2012). Other neuroimaging studies found associations between MRI abnormalities and delirium features, such as duration of delirium and subsequent cognitive impairment (for a review see Soiza et al., 2008). Recent exploratory research in intensive care unit patients suggests that duration of delirium is associated with brain atrophy, particularly superior frontal and hippocampal, leading to long-term cognitive impairment (Gunther et al., 2012). In a very similar setting, duration of delirium was positively correlated with DTI white matter tract abnormalities at discharge, which persisted at 3 months follow-up (Morandi et al., 2012).

Larger, prospective studies with high power for detecting association between baseline WMH and delirium while controlling for important covariates are lacking. In most studies sample sizes were small, and importantly without baseline pre-delirium imaging performed (Alsop et al., 2006; Soiza et al., 2008). In addition, a number of study participants had diagnosed degenerative dementias or other known structural brain diseases confounding the interpretation of imaging abnormalities (Alsop et al., 2006; Soiza et al., 2008).

Given these limitations, there remains a gap in our knowledge on the relationship between preexisting neuroanatomical changes and delirium. Our study aims to fill this gap using state-of-the-art MRI techniques within a prospective study of post-operative delirium in older persons without dementia. We investigated the association of pre-surgical MRI measures of total WMH volume, global brain atrophy and hippocampal volume with the incidence and severity of postoperative delirium. Our study extended previous work by using state-of-the-art imaging and analysis methods, cognitive and delirium assessments, in a large well-characterized clinical cohort of older individuals without dementia.

2. Methods

2.1 Study Design and Cohort Assembly

Our study population is a subsample of the Successful AGing after Elective Surgery (SAGES) study, a five-year prospective observational study of postoperative delirium being conducted at Beth Israel Deaconess Medical Center, Brigham and Women’s Hospital, and Hebrew SeniorLife in Boston (Massachusetts, USA). The study design and eligibility criteria have been reported in detail previously (Schmitt et al., 2012). Briefly, the elective surgeries included in the study were total hip or knee replacement, laminectomy, lower extremity arterial bypass, open abdominal aortic aneurysm repair, and colectomy. SAGES subjects were recruited through regular review of operating room schedules. Inclusion criteria included: age ≥ 70 years old, English speaking, and undergoing elective surgery at Beth Israel Deaconess Medical Center or Brigham and Women’s Hospital. Exclusion criteria included diagnosis of dementia as assessed by initial medical record screening, or reported by the patient during telephone recruitment or enrollment interview; cognitive impairment as defined by a score ≤ 69 or its education-adjusted equivalent on the Modified Mini-Mental State examination (3MS) during the baseline interview (Teng and Chui, 1987); terminal disease; total blindness; severe deafness; and alcohol intake > 5 drinks per day (men) or > 4 drinks per day (women). A subset of approximately one-third of the enrolled SAGES participants was recruited to undergo MRI one month prior to surgery. Additional exclusion criteria for the nested cohort MRI study included contraindications to 3 Tesla MRI, such as pacemakers and certain stents and implants. All study procedures were approved by the institutional review boards at the two surgical sites (Beth Israel Deaconess Medical Center and Brigham and Women’s Hospital) and Hebrew SeniorLife, the coordinating center. All subjects gave written informed consent.

2.2 Baseline Measures

All the study participants completed a baseline interview about two weeks before surgery. During the baseline interview, the Short Form 12 (SF-12) (Ware et al., 1996), Modified Mini-Mental State examination (3MS), complete neuropsychological test battery (described below), and the Confusion Assessment Method (CAM) (Inouye et al., 1990) were administered. We also assessed presence of depression using the Geriatrics Depression Scale (GDS), which yields a score from 0 to 15 (15 being most depressed). We dichotomized this variable at 5 with an accepted threshold of 5 or above as possibly having depression. Demographic information such as age, gender, vascular comorbidities, ethnicity, years of education and type of surgery planned (orthopedic, vascular, gastrointestinal) were collected.

2.3 Outcome Measures

The outcomes of interest were post-operative delirium incidence and severity during the hospital stay. To enhance sensitivity we utilized multiple methods to detect delirium. We used both the CAM method and chart review to detect delirium for each patient. If both, or either of the two methods indicated that the patient had delirium, then that patient was defined to have delirium. The CAM method is based on daily structured interview that includes formal neuropsychological testing. Trained interviewers assessed patients daily during hospitalization for the development of delirium using the CAM (Inouye et al., 1990), a widely used, standardized method for identification of delirium that has high sensitivity, specificity and inter-rater reliability (Wei et al., 2008). The CAM was rated based on information from patient interviews including a brief cognitive screen (e.g., orientation, short-term recall, verbal fluency, digit span) described previously (Schmitt et al., 2012), the Delirium Symptom Interview (DSI) (Albert et al., 1992), and information on mental status changes from nurses or family members. The CAM diagnostic algorithm requires the presence of acute change or fluctuating course, inattention, and either disorganized thinking or an altered level of consciousness to fulfill criteria for delirium. The validated chart review method (Inouye et al., 2005; Saczynski et al., 2014) was performed to enhance sensitivity in detecting delirium episodes that might have fallen outside the time of the CAM assessment; each chart rating was adjudicated by 2 experts. Delirium severity was assessed by the CAM-S Long Form, with scores ranging from 0 to 19 (19 = most severe) (Inouye et al., 2014b). The CAM-S provides an additive score for delirium severity based on the presence of delirium symptoms by the CAM. The CAM-S has been demonstrated to have predictive validity for clinical outcomes associated with delirium severity. Scores of 5 or greater represent severe delirium. Peak CAM-S scores were utilized in our analyses.

2.4 Covariates

The following variables were incorporated into our analysis to adjust for potential confounders: age, gender, general cognitive performance (GCP), vascular comorbidity status, and MRI derived intracranial cavity volume (ICV). Age refers to the participant’s age (years) at the time of surgery. GCP, a weighted composite to reflect the patient’s preoperative cognitive functioning, was based on a battery of neuropsychological tests to assess attention, memory, learning and executive functioning (Jones et al., 2010). The battery included the Hopkins Verbal Learning Test (Brandt, 1991), the Visual Search and Attention Task (Trenerry et al., 1990), the Trail Making Tests A and B (Trail Making Tests A and B, 1944), the Digit Symbol Substitution and Copy tests (Wechsler, 1981), the Digit Span forward and backward (Wechsler, 1981), and the 15-item Boston Naming Test (Mack et al., 1992). A higher GCP score indicates better cognitive performance. The GCP scores in SAGES were scaled to reflect population-based norms (Gross et al., 2014) with a mean value of 50 (standard deviation = 10). Thus, a score of 40 on the GCP corresponds to cognitive functioning that is one standard deviation below the mean of the U.S. population. The following conditions were considered to define vascular comorbidity: confirmed or history of myocardial infarction; congestive heart failure; peripheral vascular disease or diabetes (with or without end organ damage); cerebrovascular disease (carotid stenosis or history of stroke or transient ischemic attack); or hemiplegia. The study subjects were categorized in two categories (with vs. without vascular comorbidity), according to the presence/absence of at least one of the pathologic conditions mentioned above. Complications were assessed via chart review by two clinicians and included pneumonia, urinary tract infections, sepsis, surgical wound infection, abdominal abscess, anastomic leak, colitis, gout attack, gout flare, lactic acidosis, leukocytosis.

2.5 MRI Acquisition

All subjects were imaged at the Beth Israel Deaconess Medical Center Radiology Department on a 3 Tesla HDxt MRI (General Electric Medical Systems, Milwaukee, WI) scanner using a standard 8-channel head coil. All participants completed a standard clinical screening form for MRI safety and contraindications before the MRI scan.

The MRI acquisition protocol included:

  1. High Resolution 3D anatomic imaging – A magnetization prepared fast gradient echo (MPRAGE) 3D T1-weighted sequence with a TR of 7.9 ms, a TE of 3.2 ms, a 15° flip angle and 32 kHz bandwidth, a coronal acquisition plane with 24×19 cm field of view, 0.94 mm in plane resolution, 1.4 mm slices, a preparation time of 1100 ms with repeated saturation at the beginning of the saturation period, and an adiabatic inversion pulse 500 ms before imaging.

  2. High Resolution 3D T2-weighted imaging – A 3D fast spin echo sequence (Cube™) was acquired in the sagittal plane with TR 3 s, effective TE 63.6 ms, echo train length 100, bandwidth 62.5, 1 mm in plane resolution, and 128 1.6 mm slices.

  3. FLuid Attenuated Inversion Recovery (FLAIR) – An axial acquisition with a TR/TE of 9000/156 and an inversion time of 2250 was employed. Seventy 2 mm slices were acquired with no interslice gap and interleaved acquisition. Nominal in plane resolution was 1 mm.

2.6 MRI Analysis

Three quantitative MRI indices were extracted through image analysis for this study: 1) WMH volume, as a measure of cerebrovascular damage; 2) Brain Parenchymal Volume (BPV), as a measure of global brain atrophy; 3) hippocampal volume, as a measure of regional atrophy of the hippocampus. Deep WMH were also rated by a single, blinded rater (MC) using the visual semi-quantitative Fazekas score (Fazekas et al., 1987), ranging from 0–3 with increasing severity (0 = absence of WMH, 1 = punctate foci of WMH, 2 = early confluent WMH, 3 = confluent WMH).

Total WMH volume was obtained by a semi-automated image segmentation method based on three channels: T1-weighted, T2-weighted and FLAIR images. Bias-field correction to account for field inhomogeneities and automated spatial co-registration of T1-weighted, T2-weighted and FLAIR images was performed using the 3DSlicer software (www.slicer.org). Intracranial cavity (ICC) masks were segmented from T2-weighted images using the Brain Extraction Tool (Smith, 2002), and subsequently visually assessed and manually corrected by a single, blinded reviewer (TTH). A second rater (MC) reviewed and manually corrected ten ICV masks independently to assess inter-rater reliability. These ICC masks include the brain parenchyma as well as ventricular and cortical cerebrospinal fluid (CSF). The ICC was used as a mask to exclude non-brain/CSF voxels from the subsequent tissue class segmentation, and as intracranial cavity volume (ICV) reference to correct for head size. Tissue class segmentation of grey matter (GM), white matter (WM), and CSF from T1-weighted images occurred in a custom pipeline using FreeSurfer (version 5.2) (Fischl et al., 2002). This segmentation was corrected by use of heuristic constraints based on the FLAIR image, thereby accounting for misclassifications due to the presence of WMH. The final WMH segmentation was then obtained by a custom version of a 3-channel (T1-weighted, T2-weighted, FLAIR) Expectation Maximization (EM) algorithm, classifying lesions as outliers from the normal tissue intensity distributions (Van Leemput et al., 2001). The initial Freesurfer segmentation was used as the starting point for automated EM segmentation. The EM result is a set of probability maps that provide the likelihood for each image pixel to be part of a lesion (WMH) or one of the other tissue classes. The WMH maps were reviewed and thresholded by a neuroimaging expert (MC) to obtain binary WMH masks using FLAIR images as reference. Total WMH volume, measured using the binary WMH masks, was included in the statistical analysis.

Based on the automated output of FreeSurfer, we measured BPV and hippocampal volume as follows. No manual editing was performed at any stage of the processing workflow (Wenger et al., 2014; Mulder et al., 2014). All the brain structures segmented by FreeSurfer and included in the brain parenchyma (i.e. all of the FreeSurfer labels except the cortical and ventricular CSF, and the choroid plexi) were summed to obtain a measure of the whole BPV. Hippocampal volume measurement was based on the automated parcellation of this structure by FreeSurfer. The average volume of the left and right hippocampus was included in the statistical analysis. Post-hoc quality assessment of the hippocampal volume measures was performed by examining the degree of concordance between the left and right measurement of the hippocampal volumes within each patient. We expected the average within-patient correlation of these two measurements to be no lower than 0.60 (considered to be in the range of a good correlation). We used two measures to assess this correlation: a nonparametric Spearman correlation, and an intra-class correlation obtained from a linear mixed-effects model where the correlation was measured by the compound symmetry structure of the variance-covariance matrix of the two hippocampal volumes. We obtained a Spearman correlation of 0.69 (p-value < 0.0001), and an intra-class correlation of 0.61 (p-value < 0.0001).

Baseline T2-weighted images were not available for 14 study subjects. T2-weighted images from follow-up scans obtained at one-year following surgery were used to obtain the ICC masks and WMH probability maps in these subjects. While ICV should remain constant at the age of our study subjects, WMH measures may be perceived as biased (false positive) due to the temporal separation. However, the relative weight of the T2-channel in the workflow that classifies WMH was minimal relative to that of the FLAIR channel, and – more importantly – the final expert-driven thresholding of the resulting WMH maps was performed based on the FLAIR images of the relevant time-point. In addition, visual expert review of the 14 follow-up scans did not reveal substantial increase in WMH burden for any of the 14 subjects. The neuroimaging experts who performed the processing and analysis of the images were completely blinded to the clinical data and delirium status.

MRI volumes were all controlled for the effect of subject head size (ICV) by regression. This was achieved by including ICV as a covariate in all statistical models tested. Controlling for ICV by ratio normalization was not used as a primary approach because of potential limitations (Arndt et al., 1991; Curran-Everett, 2013; Sanfilipo et al., 2004). Parallel analyses with the ratio method were performed but no qualitatively different results were found and they are not reported here.

2.7 Statistical Analysis

Student’s t-test or Kruskal-Wallis test was used to assess differences in continuous demographic, clinical and MRI variables between subjects who developed delirium and subjects who did not, according to the distribution of the data. Chi-Square or Fisher’s exact tests were used to assess differences in categorical demographic, clinical and MRI variables between subjects who developed delirium and subjects who did not, as appropriate. Bivariate correlations between MRI indices were assessed using the nonparametric Spearman correlation coefficient (rho). Generalized linear modeling was used to estimate the predictive power of the MRI indices with respect to delirium incidence and severity. Log link and Poisson error distribution were used for the binary delirium incidence outcome (Zou, 2004). Delirium severity was analyzed with and without logarithm transformation, and the results were not different in terms of the significance of the p-values. For ease of interpretation of the model coefficients (the slopes in relations to MRI parameters), we chose to report the non-transformed scale of delirium severity. For each MRI variable, both unadjusted and adjusted analyses were conducted. For unadjusted analyses, only ICV was included in the model to control for head size effects according to the normalization by regression method. For adjusted analyses, the following covariates were included in the model: ICV, age, gender, GCP and vascular comorbidity. According to formal power analysis prior to the study, the available sample of 146 provided a power of at least 0.80 to detect a population odds ratio of 2.0 per one standard deviation difference in hippocampal volume under the assumption that the incidence of delirium in this sample was 20%, and the association in terms of odds ratio between delirium incidence and each of the four covariates (age, GCP, gender, vascular comorbidity) in the model was at least 1.05. Complications were coded as binary (presence/absence) and were included in the regression models in post-hoc analysis to investigate first if there was a significant effect on delirium incidence or severity, and second if the MRI effect estimates would change in the presence/absence of complications. As a test of MRI data quality, generalized linear modeling was also performed to assess the dependence of MRI variables on these covariates for comparison with prior studies. Type-I was set to 0.017 (a Bonferroni correction of 0.05 for the 3 MRI variables tested). SAS software version 9.3 (SAS/STAT User’s Guide, Version 8, SAS Institute, Cary, N.C., 2000) was used for all statistical analyses in this paper.

3. Results

Of 147 subjects who had MRI prior to surgery, 1 did not complete the entire MRI protocol, and therefore was excluded from this study. Demographic and baseline clinical characteristics of the 146 subjects included in this study are summarized in Table 1. Of 146 subjects, 32 (22%) developed delirium during hospitalization. Delirium was diagnosed in 23 subjects by both the CAM interview and chart review method, and in 9 subjects by chart review alone. No significant differences were found between subjects with and without delirium in age, gender, or 3MS (Table 1). Subjects with delirium had lower GCP (55 ± 7 vs. 59 ± 7, p = 0.002, Student’s t-test), and lower SF-12 Physical composite scores (32 ± 10 vs. 37 ± 10, p = 0.016, Student’s t-test). Subjects with delirium consistently scored higher on delirium severity rating as compared to those without delirium (Table 1). We found no statistically significant difference in delirium incidence between subjects with and without depression. Proportion of delirium was 19% (4/21) in subjects with depression, and 22% (28/125) in those without depression.

Table 1.

Baseline characteristics of study participants

All Subjects No Delirium Delirium p-value
Number of Subjects (n, %) 146 114 32 -
Age (years, mean ± SD) 76 ± 4 76 ± 5 77 ± 4 0.467a
Female Gender (n, %) 87 (60%) 65 (57%) 22 (69%) 0.232c
Non-white or Hispanic (n, %) 14 (10%) 11 (10%) 3 (9%) 1.000d
Education (years, mean ± SD) 15 ± 3 15 ± 3 14 ± 3 0.188a
3MS Score (0–30, 0 most severe; mean ± SD) 26 ± 2 26 ± 1 26 ± 2 0.159a
GCP Score (externally scaled, mean ± SD) 58 ± 7 59 ± 7 55 ± 7 0.002a
SF-12 Physical Composite (T-score, mean ± SD) 36 ± 10 37 ± 10 32 ± 10 0.016a
Vascular Comorbidity (n, %) 58 (40%) 45 (39%) 13 (41%) 0.906c
    • Cardiovascular disease 28 (19%) 18 (16%) 10 (31%) 0.049c
    • Congestive Heart Failure 3 (2%) 2 (2%) 1 (3%) 0.527d
    • Peripheral vascular disease 11 (8%) 8 (7%) 3 (9%) 0.706d
    • Hemiplegia 0 - - -
    • Diabetes 28 (19%) 20 (18%) 8 (25%) 0.344c
    • Diabetes w/end organ damage 5 (3%) 4 (4%) 1 (3%) 1.000d
Surgery (n, %)
    • Orthopedic 120 (82%) 92 (81%) 28 (88%) 0.782d
    • Vascular 8 (5%) 7 (6%) 1 (3%)
    • Gastrointestinal 18 (12%) 15 (13%) 3 (9%)
Delirium Severity (mean ± SD)
    • CAM-S Long (0–19, 19 most severe) 3.59 ± 3.2 2.37 ± 1.6 7.94 ± 3.8 <0.001b
Fazekas Score (0–3, 3 most severe; n, %)
    • 0 10 (7%) 9 (8%) 1 (3%) 0.358d
    • 1 97 (66%) 72 (63%) 25 (78%)
    • 2 39 (27%) 33 (29%) 6 (19%)
    • 3 0 - -
Open in a new tab

P-values refer to group comparison no delirium vs. delirium by (a) Student’s t-test, (b) Kruskal-Wallis test, (c) Chi-Squared test or (d) Fisher’s exact test.

Abbreviations: CAM-S – Confusion Assessment Method - Severity; GCP – Global Cognitive Performance; MDAS – Memorial Delirium Assessment Scale; 3MS – Modified Mini-Mental State Examination; SF-12 – 12-item Short Form.

WMH volume, BPV and hippocampal volume properties are summarized in Table 2. BPV and hippocampal volume were normally distributed, whereas WMH volume showed kurtosis and skewness. BPV showed a significant association with hippocampal volume (rho = 0.62, p < 0.001, Spearman’s correlation). As expected, WMH volume correlated with Fazekas score (rho = 0.71, p < 0.001, Spearman’s correlation). WMH volume did not show significant association with BPV or hippocampal volume (Spearman’s correlation). GCP showed no significant correlation with BPV, hippocampal volume and WMH volume (Spearman’s correlation). Generalized linear modeling of MRI variables showed significant dependence on age with slopes −5.15 ml/year (p < 0.001), −0.02 ml/year (p = 0.004), and 0.52 ml/year (p = 0.003) for BPV, average hippocampal volume, and WMH volume, respectively. Significant dependence of 5.9 ml greater WMH volume for females was also found (p = 0.004). Coefficients of variation of the residuals were 6% for BPV, 11% for hippocampal volume, and 78% for WMH volume.

Table 2.

Quantitative pre-surgical MRI measures of brain parenchymal damage

All Subjects (n=146) No Delirium (n=114) Delirium (n=32) p-value
WMH Volume (cc) 11.27 ± 9.46 11.55 ± 9.94 10.24 ± 7.59 0.710b
BPV (cc) 1013.91 ± 113.11 1018.71 ± 114.32 996.79 ± 108.68 0.334a
Hippocampal Volume (cc) 3.24 ± 0.43 3.23 ± 0.43 3.25 ± 0.47 0.862a
ICV (cc) 1416.71 ± 158.51 1417.88 ± 163.73 1410.05 ± 138.23 0.805a
Open in a new tab

Data are expressed as mean ± SD.

P-values refer to group comparison no delirium vs. delirium by (a) Student’s t-test or (b) Kruskal-Wallis test.

Abbreviations: BPV – Brain Parenchymal Volume; ICV – Intracranial Cavity Volume; WMH – White Matter Hyperintensities.

Differences in WMH volume, BPV and hippocampal volume between subjects who developed later delirium during their hospital stay and subjects who did not were not statistically significant (Table 2). Both unadjusted and adjusted generalized linear model analyses demonstrated no statistically significant effects for any of the MRI indices with respect to delirium incidence or severity as measured by the CAM-S Long Form (Tables 3 and 4). GCP score was a significant covariate of delirium incidence (p < 0.01) and severity (p < 0.001) in both multivariable models. The prevalence of complications was 13% (19/146). The main effect of complications was not statistically significant for delirium incidence (p = 0.874 for BPV; p = 0.789 for hippocampal volume; p = 0.808 for WMH; generalized linear model), and significant for delirium severity (p = 0.020 for BPV; p = 0.019 for hippocampal volume; p = 0.014 for WMH; general linear model). The estimates of the MRI parameters in the presence of the significant effect of complications showed little changes, and still remained not statistically significant.

Table 3.

Effect of pre-surgical MRI indices of brain damage on postoperative delirium incidence

Relative Risk
(per cc unit change)		 95%-C.I. p-value
WMH Volume
Adjusted for ICV only 0.987 0.947 – 1.029 0.549
Fully adjusted model* 0.966 0.919 – 1.016 0.178
BPV
Adjusted for ICV only 0.997 0.992 – 1.002 0.251
Fully adjusted model* 0.997 0.992 – 1.003 0.367
Hippocampal Volume
Adjusted for ICV only 1.160 0.453 – 2.973 0.757
Fully adjusted model* 1.371 0.494 – 3.799 0.545
Open in a new tab

Generalized linear model with log link and Poisson error distribution.

*

The fully adjusted model includes the following covariates: ICV, age, gender, global cognitive performance and vascular comorbidity.

Abbreviations: BPV – Brain Parenchymal Volume; ICV – Intracranial Cavity Volume; WMH – White Matter Hyperintensities.

Table 4.

Effect of pre-surgical MRI indices of brain damage on delirium severity as assessed by the Confusion Assessment Method - Severity test Long Form

Estimate
(per cc unit change)		 95%-C.I. p-value
WMH Volume
Adjusted for ICV only −0.032 −0.087 – 0.024 0.262
Fully adjusted model* −0.055 −0.107 – −0.002 0.045
BPV
Adjusted for ICV only −0.006 −0.014 – 0.002 0.136
Fully adjusted model* −0.002 −0.009 – 0.006 0.686
Hippocampal Volume
Adjusted for ICV only −0.494 −1.892 – 0.904 0.490
Fully adjusted model* 0.178 −1.160 – 1.515 0.795
Open in a new tab

General linear model.

*

The fully adjusted model includes the following covariates: ICV, age, gender, global cognitive performance and vascular comorbidity.

Abbreviations: BPV – Brain Parenchymal Volume; ICV – Intracranial Cavity Volume; WMH – White Matter Hyperintensities.

4. Discussion

We investigated the predictive role of MRI-derived measures of brain damage with respect to delirium incidence and severity within a prospective study of older subjects without dementia undergoing elective surgery. Incidence of delirium in our study sample (22%) is in line with data reported in the literature (Inouye et al., 2014a). We used WMH volume, BPV and hippocampal volume as established MRI indicators of cerebrovascular damage, global and regional brain atrophy, respectively. None of these three MRI indices of brain damage was significantly associated with delirium incidence or severity in both unadjusted and adjusted analysis in our study.

This finding appears to be in contrast with recent studies that showed MRI correlates of white matter damage and brain atrophy in subjects with delirium. However, most of these studies had methodological limitations inherent in studying the difficult problem of delirium (Alsop et al., 2006; Soiza et al., 2008), such as a retrospective study design (Hatano et al., 2013; Root et al., 2013), lack of a control group or no baseline pre-delirium imaging (Gunther et al., 2012; Morandi et al., 2012). Recently, a large prospective study (153 subjects aged 60 years or older, of which 16 developed delirium) reported a positive association between multiple brain infarcts on pre-surgical MRI and delirium occurrence after cardiac surgery (Otomo et al., 2013). However, the frequency of pure existing single brain infarcts was higher in the non-delirium group, and there was no association between incidence of delirium and WMH burden as measured by the Fazekas score. A number of the studies were also underpowered to establish a robust cause-effect link between preexisting brain pathology and delirium (Figiel et al., 1990a, 1990b). In some of these studies, subjects with underlying neurological or neurodegenerative diseases that may contribute to confounding MRI abnormalities were included (Alsop et al., 2006; Otomo et al., 2013; Soiza et al., 2008). Furthermore, the heterogeneity of underlying medical conditions in the cohorts examined – surgery, critical care (e.g., sepsis, respiratory failure, mechanical ventilation), and general medical illness – may limit the generalizability of the neuroimaging findings in these previous studies. A DTI study showed that microstructural abnormalities of the cerebral white matter and thalamus, as detected by pre-surgical MRI, predicted occurrence of delirium in 19 of 116 elective cardiac surgery patients (Shioiri et al., 2010). Similar to our study, the authors mentioned (without reporting the results) no quantitative differences in MRI measures derived from T1- and T2-wighted images between subjects with and without delirium. Although DTI might be more sensitive than WMH volume to detect subtle structural abnormalities, analysis of the difference in diffusion parameters were not adjusted for relevant potential confounders showing higher prevalence in the delirium group, such as hypertension, hemodialysis and blood transfusion.

The strength and validity of our results lie in the prospective study design and robust means of pre-evaluation and follow-up of a study population that is part of the larger SAGES study (Schmitt et al., 2012). Our study has the largest sample size yet investigated to study the relationship between structural brain damage and delirium using state-of-the-art MRI indices of neurodegenerative and cerebrovascular injury.

Accuracy of the MRI-derived measures employed in this study is further supported by our secondary findings of association of age with WMH volume, GCP and BPV, as well as residual coefficients of variation that are consistent with previous literature (Atwood et al., 2004; Jack et al., 1992; Whitwell et al., 2001; Coffey et al., 1992; Convit et al., 1995; Poggesi et al., 2011; Wu et al., 1981). The observed significant association between GCP score and delirium incidence and severity is also in line with previous literature, which shows that lower cognitive function is among the leading predisposing factors for delirium (Inouye, 2006; Inouye et al., 2014a, 1993; Marcantonio et al., 1994).

Several important limitations of the study are worthy of comment. Since subjects with clinically evident dementia were excluded, the generalizability of our findings to this important population of subjects requires further study. Moreover, since this study was based on an elective surgical population in a single geographic area, future studies will be needed to verify the findings in other settings and populations, such as general medical, intensive care, and post-acute settings.

Although our findings do not support the utility of MRI-derived indices of global WMH volume and whole-brain or hippocampal atrophy for predicting delirium, we cannot rule out a potential role for regional WMH burden, volumes of specific brain regions or other measures of brain pathology with regards to delirium incidence. This study examined only the role of pre-delirium (baseline) neuroimaging markers for the prediction of delirium; examination of the impact of delirium on subsequent neuroimaging measures could not be examined in the present study. MRI correlates of brain damage may still prove to be useful in monitoring subjects following delirium. For example, quantitative MRI indices of cerebral white matter damage and atrophy may have prognostic value for predicting and monitoring disabling sequelae of delirium, such as cognitive decline. In addition, interval change in these volumetric brain indices may have significant correlation with delirium incidence or severity. These represent important areas for future research.

Delirium is a morbid yet common condition, with significant impact on the outcome and long-term prognosis of older patients. This study raises the intriguing possibility that delirium incidence in elective surgery patients may not be strongly related to underlying generalized or hippocampal atrophy, or WMH, at least in a population without dementia at baseline.

Supplementary Material

NIHMS671513-supplement.docx (104.2KB, docx)

Highlights.

  • Prospective study of the association between MRI and postoperative delirium

  • Powered study of delirium using state-of-the-art MRI indices of structural brain damage

  • Pre-surgical white matter hyperintensities, brain and hippocampal volume were measured

  • Out of 146 consecutively enrolled subjects, 32 (22%) developed delirium

  • None of the three MRI indices was associated with delirium incidence or severity

Acknowledgements

A list of participating personnel of the SAGES Study can be found online as supplementary material.

Funding

Supported by Grants No. P01AG031720 (SKI), K07AG041835 (SKI), and K24AG035075 (ERM) from the National Institute on Aging. Dr. Hshieh is supported by a NIH funded T32 Training Grant (AG000158). Dr. Inouye holds the Milton and Shirley F. Levy Family Chair. The funding sources had no role in the design, conduct, or reporting of this study.

Abbreviations

(BPV)

Brain Parenchymal Volume

(CSF)

Cerebrospinal Fluid

(CAM)

Confusion Assessment Method

(DSI)

Delirium Symptom Interview

(DTI)

Diffusion Tensor Imaging

(FLAIR)

Fluid Attenuated Inversion Recovery

(GCP)

General Cognitive Performance

(ICC)

Intracranial Cavity

(ICV)

Intracranial Volume

(MRI)

Magnetic Resonance Imaging

(MPRAGE)

Magnetization Prepared Fast Gradient Echo

(MDAS)

Memorial Delirium Assessment Scale

(3MS)

Modified Mini-Mental State Examination

(SF-12)

Short Form 12

(WMH)

White Matter Hyperintensity

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Disclosure

All authors report no disclosure.

References

  1. Albert MS, Levkoff SE, Reilly C, Liptzin B, Pilgrim D, Cleary PD, Evans D, Rowe JW. The delirium symptom interview: an interview for the detection of delirium symptoms in hospitalized patients. J. Geriatr. Psychiatry Neurol. 1992;5:14–21. doi: 10.1177/002383099200500103. [DOI] [PubMed] [Google Scholar]
  2. Alsop DC, Fearing Ma, Johnson K, Sperling R, Fong TG, Inouye SK. The role of neuroimaging in elucidating delirium pathophysiology. J. Gerontol. A. Biol. Sci. Med. Sci. 2006;61:1287–1293. doi: 10.1093/gerona/61.12.1287. [DOI] [PubMed] [Google Scholar]
  3. Arndt S, Cohen G, Alliger RJ, Swayze VW, Andreasen NC. Problems with ratio and proportion measures of imaged cerebral structures. Psychiatry Res. 1991;40:79–89. doi: 10.1016/0925-4927(91)90031-k. [DOI] [PubMed] [Google Scholar]
  4. Atwood LD, Wolf PA, Heard-Costa NL, Massaro JM, Beiser A, D’Agostino RB, DeCarli C. Genetic variation in white matter hyperintensity volume in the Framingham Study. Stroke. 2004;35:1609–1613. doi: 10.1161/01.STR.0000129643.77045.10. [DOI] [PubMed] [Google Scholar]
  5. Brandt J. The hopkins verbal learning test: Development of a new memory test with six equivalent forms. Clin. Neuropsychol. 1991;5:125–142. [Google Scholar]
  6. Choi S-H, Lee H, Chung T-S, Park K-M, Jung Y-C, Kim SI, Kim J-J. Neural network functional connectivity during and after an episode of delirium. Am. J. Psychiatry. 2012;169:498–507. doi: 10.1176/appi.ajp.2012.11060976. [DOI] [PubMed] [Google Scholar]
  7. Coffey CE, Wilkinson WE, Parashos IA, Soady SA, Sullivan RJ, Patterson LJ, Figiel GS, Webb MC, Spritzer CE, Djang WT. Quantitative cerebral anatomy of the aging human brain: a cross-sectional study using magnetic resonance imaging. Neurology. 1992;42:527–536. doi: 10.1212/wnl.42.3.527. [DOI] [PubMed] [Google Scholar]
  8. Convit A, de Leon MJ, Tarshish C, De Santi S, Kluger A, Rusinek H, George AE. Hippocampal volume losses in minimally impaired elderly. Lancet. 1995;345:266. doi: 10.1016/s0140-6736(95)90265-1. [DOI] [PubMed] [Google Scholar]
  9. Curran-Everett D. Explorations in statistics: the analysis of ratios and normalized data. Adv. Physiol. Educ. 2013;37:213–219. doi: 10.1152/advan.00053.2013. [DOI] [PubMed] [Google Scholar]
  10. Fazekas F, Chawluk JB, Zimmerma A, June M. MR Signal Abnormalities at 1.5 T in Alzheimer’s Dementia and Normal Aging deficiency. AJR. Am. J. Roentgentol. 1987;149:351–356. doi: 10.2214/ajr.149.2.351. [DOI] [PubMed] [Google Scholar]
  11. Figiel GS, Coffey CE, Djang WT, Hoffman G, Doraiswamy PM. Brain magnetic resonance imaging findings in ECT-induced delirium. J. Neuropsychiatry Clin. Neurosci. 1990;2:53–58. doi: 10.1176/jnp.2.1.53. [DOI] [PubMed] [Google Scholar]
  12. Figiel GS, Krishnan KR, Doraiswamy PM. Subcortical structural changes in ECT-induced delirium. J. Geriatr. Psychiatry Neurol. 1990;3:172–176. doi: 10.1177/089198879000300309. [DOI] [PubMed] [Google Scholar]
  13. Fischl B, Salat DH, Busa E, Albert M, Dieterich M, Haselgrove C, van der Kouwe A, Killiany R, Kennedy D, Klaveness S, Montillo A, Makris N, Rosen B, Dale AM. Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron. 2002;33:341–355. doi: 10.1016/s0896-6273(02)00569-x. [DOI] [PubMed] [Google Scholar]
  14. Fong TG, Bogardus ST, Daftary A, Auerbach E, Blumenfeld H, Modur S, Leo-Summers L, Seibyl J, Inouye SK. Cerebral perfusion changes in older delirious patients using 99mTc HMPAO SPECT. J. Gerontol. A. Biol. Sci. Med. Sci. 2006;61:1294–1299. doi: 10.1093/gerona/61.12.1294. [DOI] [PubMed] [Google Scholar]
  15. Gross AL, Jones RN, Fong TG, Tommet D, Inouye SK. Calibration and Validation of an Innovative Approach for Estimating General Cognitive Performance. Neuroepidemiology. 2014;42:144–153. doi: 10.1159/000357647. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gunther ML, Morandi A, Krauskopf E, Pandharipande P, Girard TD, Jackson JC, Thompson J, Shintani AK, Geevarghese S, Miller RR, Canonico A, Merkle K, Cannistraci CJ, Rogers BP, Gatenby JC, Heckers S, Gore JC, Hopkins RO, Ely EW. The association between brain volumes, delirium duration, and cognitive outcomes in intensive care unit survivors: the VISIONS cohort magnetic resonance imaging study. Crit. Care Med. 2012;40:2022–2032. doi: 10.1097/CCM.0b013e318250acc0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hatano Y, Narumoto J, Shibata K, Matsuoka T, Taniguchi S, Hata Y, Yamada K, Yaku H, Fukui K. White-matter hyperintensities predict delirium after cardiac surgery. Am. J. Geriatr. Psychiatry. 2013;21:938–945. doi: 10.1016/j.jagp.2013.01.061. [DOI] [PubMed] [Google Scholar]
  18. Hughes CG, Patel MB, Pandharipande PP. Pathophysiology of acute brain dysfunction: what’s the cause of all this confusion? Curr. Opin. Crit. Care. 2012;18:518–526. doi: 10.1097/MCC.0b013e328357effa. [DOI] [PubMed] [Google Scholar]
  19. Inouye SK. Delirium in older persons. N. Engl. J. Med. 2006;354:1157–1165. doi: 10.1056/NEJMra052321. [DOI] [PubMed] [Google Scholar]
  20. Inouye SK, Leo-Summers L, Zhang Y, Bogardus ST, Leslie DL, Agostini JV. A chart-based method for identification of delirium: validation compared with interviewer ratings using the confusion assessment method. J. Am. Geriatr. Soc. 2005;53:312–318. doi: 10.1111/j.1532-5415.2005.53120.x. [DOI] [PubMed] [Google Scholar]
  21. Inouye SK, van Dyck CH, Alessi CA, Balkin S, Siegal AP, Horwitz RI. Clarifying confusion: the confusion assessment method. A new method for detection of delirium. Ann. Intern. Med. 1990;113:941–948. doi: 10.7326/0003-4819-113-12-941. [DOI] [PubMed] [Google Scholar]
  22. Inouye SK, Viscoli CM, Horwitz RI, Hurst LD, Tinetti ME. A predictive model for delirium in hospitalized elderly medical patients based on admission characteristics. Ann. Intern. Med. 1993;119:474–481. doi: 10.7326/0003-4819-119-6-199309150-00005. [DOI] [PubMed] [Google Scholar]
  23. Inouye SK, Westendorp RGJ, Saczynski JS. Delirium in elderly people. Lancet. 2014a;383:911–922. doi: 10.1016/S0140-6736(13)60688-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Inouye SK, Kosar CM, Tommet D, Schmitt EM, Puelle MR, Saczynski JS, Marcantonio ER, Jones RN. The CAM-S: Development and Validation of a New Scoring System for Delirium Severity in 2 Cohorts. Ann. Intern. Med. 2014b;160:526–533. doi: 10.7326/M13-1927. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Jack CR, Petersen RC, O’Brien PC, Tangalos EG. MR-based hippocampal volumetry in the diagnosis of Alzheimer’s disease. Neurology. 1992;42:183–188. doi: 10.1212/wnl.42.1.183. [DOI] [PubMed] [Google Scholar]
  26. Jones RN, Rudolph JL, Inouye SK, Yang FM, Fong TG, Milberg WP, Tommet D, Metzger ED, Cupples LA, Marcantonio ER. Development of a unidimensional composite measure of neuropsychological functioning in older cardiac surgery patients with good measurement precision. J. Clin. Exp. Neuropsychol. 2010;32:1041–1049. doi: 10.1080/13803391003662728. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Koponen H, Hurri L, Stenbäck U, Mattila E, Soininen H, Riekkinen PJ. Computed tomography findings in delirium. J. Nerv. Ment. Dis. 1989;177:226–231. doi: 10.1097/00005053-198904000-00006. [DOI] [PubMed] [Google Scholar]
  28. Koponen H, Hurri L, Stenbäck U, Riekkinen PJ. Acute confusional states in the elderly: a radiological evaluation. Acta Psychiatr. Scand. 1987;76:726–731. doi: 10.1111/j.1600-0447.1987.tb02946.x. [DOI] [PubMed] [Google Scholar]
  29. Lipowski ZJ. Organic mental disorders and DSM-IV. Am. J. Psychiatry. 1990;147:947. doi: 10.1176/ajp.147.7.947a. [DOI] [PubMed] [Google Scholar]
  30. Mack WJ, Freed DM, Williams BW, Henderson VW. Boston Naming Test: shortened versions for use in Alzheimer’s disease. J. Gerontol. 1992;47:P154–P158. doi: 10.1093/geronj/47.3.p154. [DOI] [PubMed] [Google Scholar]
  31. Marcantonio ER, Goldman L, Mangione CM, Ludwig LE, Muraca B, Haslauer CM, Donaldson MC, Whittemore AD, Sugarbaker DJ, Poss R. A clinical prediction rule for delirium after elective noncardiac surgery. JAMA. 1994;271:134–139. [PubMed] [Google Scholar]
  32. Morandi A, Rogers BP, Gunther ML, Merkle K, Pandharipande P, Girard TD, Jackson JC, Thompson J, Shintani AK, Geevarghese S, Miller RR, Canonico A, Cannistraci CJ, Gore JC, Ely EW, Hopkins RO. The relationship between delirium duration, white matter integrity, and cognitive impairment in intensive care unit survivors as determined by diffusion tensor imaging: the VISIONS prospective cohort magnetic resonance imaging study. Crit. Care Med. 2012;40:2182–2189. doi: 10.1097/CCM.0b013e318250acdc. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Mulder ER, de Jong RA, Knol DL, van Schijndel RA, Cover KS, Visser PJ, Barkhof F, Vrenken H Alzheimer's Disease Neuroimaging Initiative. Hippocampal volume change measurement: quantitative assessment of the reproducibility of expert manual outlining and the automated methods FreeSurfer and FIRST. Neuroimage. 2014;15:169–181. doi: 10.1016/j.neuroimage.2014.01.058. [DOI] [PubMed] [Google Scholar]
  34. Otomo S, Maekawa K, Goto T, Baba T, Yoshitake A. Pre-existing cerebral infarcts as a risk factor for delirium after coronary artery bypass graft surgery. Interact. Cardiovasc. Thorac. Surg. 2013;17:799–804. doi: 10.1093/icvts/ivt304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Poggesi A, Pantoni L, Inzitari D, Fazekas F, Ferro J, O’Brien J, Hennerici M, Scheltens P, Erkinjuntti T, Visser M, Langhorne P, Chabriat H, Waldemar G, Wallin A, Wahlund A. 2001–2011: A Decade of the LADIS (Leukoaraiosis And DISability) Study: What Have We Learned about White Matter Changes and Small-Vessel Disease? Cerebrovasc. Dis. 2011;32:577–588. doi: 10.1159/000334498. [DOI] [PubMed] [Google Scholar]
  36. Root JC, Pryor KO, Downey R, Alici Y, Davis ML, Holodny A, Korc-Grodzicki B, Ahles T. Association of pre-operative brain pathology with post-operative delirium in a cohort of non-small cell lung cancer patients undergoing surgical resection. Psychooncology. 2013;22:2087–2094. doi: 10.1002/pon.3262. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Saczynski JS, Kosar CM, Xu G, Puelle MR, Schmitt E, Jones RN, Marcantonio ER, Wong B, Isaza I, Inouye SK. A tale of two methods: chart and interview methods for identifying delirium. J. Am. Geriatr. Soc. 2014;62:518–524. doi: 10.1111/jgs.12684. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Saczynski JS, Marcantonio ER, Quach L, Fong TG, Gross A, Inouye SK, Jones RN. Cognitive trajectories after postoperative delirium. N. Engl. J. Med. 2012;367:30–39. doi: 10.1056/NEJMoa1112923. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Sanfilipo MP, Benedict RHB, Zivadinov R, Bakshi R. Correction for intracranial volume in analysis of whole brain atrophy in multiple sclerosis: the proportion vs. residual method. Neuroimage. 2004;22:1732–1743. doi: 10.1016/j.neuroimage.2004.03.037. [DOI] [PubMed] [Google Scholar]
  40. Schmitt EM, Marcantonio ER, Alsop DC, Jones RN, Rogers SO, Fong TG, Metzger E, Inouye SK. Novel risk markers and long-term outcomes of delirium: the successful aging after elective surgery (SAGES) study design and methods. J. Am. Med. Dir. Assoc. 2012;13 doi: 10.1016/j.jamda.2012.08.004. 818.e1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Shioiri A, Kurumaji A, Takeuchi T, Matsuda H, Arai H, Nishikawa T. White matter abnormalities as a risk factor for postoperative delirium revealed by diffusion tensor imaging. Am. J. Geriatr. Psychiatry. 2010;18:743–753. doi: 10.1097/JGP.0b013e3181d145c5. [DOI] [PubMed] [Google Scholar]
  42. Siepe M, Pfeiffer T, Gieringer A, Zemann S, Benk C, Schlensak C, Beyersdorf F. Increased systemic perfusion pressure during cardiopulmonary bypass is associated with less early postoperative cognitive dysfunction and delirium. Eur. J. Cardiothorac. Surg. 2011;40:200–207. doi: 10.1016/j.ejcts.2010.11.024. [DOI] [PubMed] [Google Scholar]
  43. Smith SM. Fast robust automated brain extraction. Hum. Brain Mapp. 2002;17:143–155. doi: 10.1002/hbm.10062. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Soiza RL, Sharma V, Ferguson K, Shenkin SD, Seymour DG, Maclullich AMJ. Neuroimaging studies of delirium: a systematic review. J. Psychosom. Res. 2008;65:239–248. doi: 10.1016/j.jpsychores.2008.05.021. [DOI] [PubMed] [Google Scholar]
  45. Teng EL, Chui HC. The Modified Mini-Mental State (3MS) examination. J. Clin. Psychiatry. 1987;48:314–318. [PubMed] [Google Scholar]
  46. Trail Making Tests A and B. War Department, Adjutant General’s Office. Washington, DC: 1944. [Google Scholar]
  47. Trenerry M, Crosson B, DeBoe J, Leber W. Visual Search and Attention Test (VSAT) Odessa, FL: Psychological Assessment Resources; 1990. [Google Scholar]
  48. Trzepacz PT. Is there a final common neural pathway in delirium? Focus on acetylcholine and dopamine. Semin. Clin. Neuropsychiatry. 2000;5:132–148. doi: 10.153/SCNP00500132. [DOI] [PubMed] [Google Scholar]
  49. Van Leemput K, Maes F, Vandermeulen D, Colchester a, Suetens P. Automated segmentation of multiple sclerosis lesions by model outlier detection. IEEE Trans. Med. Imaging. 2001;20:677–688. doi: 10.1109/42.938237. [DOI] [PubMed] [Google Scholar]
  50. Wenger E, Mårtensson J, Noack H, Bodammer NC, Kühn S, Schaefer S, Heinze HJ, Düzel E, Bäckman L, Lindenberger U, Lövdén M. Comparing manual and automatic segmentation of hippocampal volumes: reliability and validity issues in younger and older brains. Hum Brain Mapp. 2014;35:4236–4248. doi: 10.1002/hbm.22473. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Ware J, Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Med. Care. 1996;34:220–233. doi: 10.1097/00005650-199603000-00003. [DOI] [PubMed] [Google Scholar]
  52. Wechsler D. Manual: Wechsler Adult Intelligence Scale - Revised. New York: Psychological Corp; 1981. [Google Scholar]
  53. Wei LA, Fearing MA, Sternberg EJ, Inouye SK. The Confusion Assessment Method: a systematic review of current usage. J. Am. Geriatr. Soc. 2008;56:823–830. doi: 10.1111/j.1532-5415.2008.01674.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Whitwell JL, Crum WR, Watt HC, Fox NC. Normalization of cerebral volumes by use of intracranial volume: implications for longitudinal quantitative MR imaging. AJNR. Am. J. Neuroradiol. 2001;22:1483–1489. [PMC free article] [PubMed] [Google Scholar]
  55. Witlox J, Eurelings LSM, de Jonghe JFM, Kalisvaart KJ, Eikelenboom P, van Gool WA. Delirium in elderly patients and the risk of postdischarge mortality, institutionalization, and dementia: a meta-analysis. JAMA. 2010;304:443–451. doi: 10.1001/jama.2010.1013. [DOI] [PubMed] [Google Scholar]
  56. Wu S, Schenkenberg T, Wing SD, Osborn AG. Cognitive correlates of diffuse cerebral atrophy determined by computed tomography. Neurology. 1981;31:1180–1184. doi: 10.1212/wnl.31.9.1180. [DOI] [PubMed] [Google Scholar]
  57. Zou G. A modified poisson regression approach to prospective studies with binary data. Am. J. Epidemiol. 2004;159:702–706. doi: 10.1093/aje/kwh090. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

NIHMS671513-supplement.docx (104.2KB, docx)

RESOURCES