A longitudinal investigation of locus coeruleus norepinephrine system integrity and cognitive performance across the Alzheimer’s spectrum

Abstract

Loss of locus coeruleus (LC) integrity, a debilitating feature of Alzheimer’s disease (AD), can be visualized in vivo with MRI. Longitudinal investigation across the AD spectrum can elucidate the timing of LC damage and its behavioral correlates. Cognitively normal adults (n = 153), individuals with mild cognitive impairment (n = 73), and with AD (n = 43) underwent MR imaging to assess LC integrity, PET imaging to assess tau and Aβ ([¹⁸F]MK6240 and [¹⁸F]AZD4694, respectively) and completed a battery of cognitive tests. For a subset of participants, these assessments were repeated annually for up to 4 years. Linear mixed effects analysis examined the association of time and cognitive performance with LC signal. Reduction in LC integrity over time was significant for patients early in the course of AD (t₉₅ = −3.72, p = 0.0003) but not in later AD phases or healthy aging. In AD participants, several cognitive measures correlated with LC integrity including logical memory (t₆₄ = 2.57, p = 0.013) and object recognition (t₆₄ = 3.41, p = 0.001). LC imaging reveals degeneration of the LC-norepinephrine system within subjects over time in early AD phases. LC damage correlates to cognitive impairment and the strong relationship observed of LC signal to object recognition could suggest an important role of this system in novelty detection in AD.

Keywords

Aging Alzheimer’s disease cognition locus coeruleus norepinephrine

Introduction

The locus coeruleus (LC) is the primary source of norepinephrine (NE) in the brain. This neuromodulator exerts widespread influence over most brain cortical and subcortical regions.^1–4 NE released from the LC during arousing, mentally challenging, or novel stimuli supports several cognitive processes including episodic memory, working memory, and cognitive control.^5,6 Neurodegeneration and dysfunction of the LC–NE system occurs in Alzheimer’s disease (AD) and other neurodegenerative disorders^7–11 and is associated with cognitive deficits related to learning, memory, and decision making.^5,6,12 The LC may be the first brain structure to begin accumulating hyperphosphorylated tau, even in some cases decades prior to onset of AD and may propagate the spread of tau in the brain.^13–15 Later in the illness course, this is succeeded by alterations in LC structure and function and ultimately degeneration of LC neurons.^16–22 LC–NE system dysregulation also disrupts the important role of this system in neuroprotection.²³ The consequences of these insults to the LC–NE system on the pathophysiology and symptoms of AD are not adequately understood and more work is needed to disentangle the role of the LC within the cascade of other pathological processes.

To support this line of research, recent studies have employed in vivo imaging of the LC using a method sometimes referred to as neuromelanin-sensitive MRI (NM–MRI) that provides a reproducible and practical method to assay LC integrity.^24,25 Reduced LC signal, reflecting degeneration of norepinephrine neurons in this structure^26,27 is observed in Alzheimer’s disease (AD)^28–30 where it is correlated to other pathophysiological markers, including cortical atrophy and amyloid (Aβ) and tau burden throughout the brain.^31,32 In individuals with mild cognitive impairment (MCI), reduced LC signal is also observed,³³ especially in those who convert to AD.^34,35 However, inferring LC signal loss on the basis of low LC signal in cross-sectional data is suboptimal and there has been little longitudinal work to investigate LC signal loss within individuals over time. This would support more accurate identification of when LC signal loss begins, when it is most pronounced, and what are its behavioral consequences, thereby elucidating its potential as an early marker of AD risk.

Given the important role of the LC–NE system in cognition, several studies have examined the correlation of LC signal loss to cognitive impairment in cognitively normal older adults, as well as MCI and AD patients.^31,36 In work combining older adults with and without MCI, associations were found between reduced rostral/middle LC–CNR and several cognitive domains: verbal fluency, episodic memory, and processing speed with these relationships driven more strongly by individuals with MCI.^31,37 In healthy aging, some studies have found LC signal to be positively associated with global cognition,³⁸ episodic memory,³⁹ and other specific cognitive functions (e.g. emotional regulation, inhibitory control, and emotional memory),³⁸ while other studies in cognitively normal older adults found no correlations between LC signal and cognition^35,40 consistent with the possibility that variation in LC signal in the absence of degeneration has minimal effects on cognition. Differences between studies in the strictness of the definition of “healthy aging” could account for some of these discrepancies. Consistent with this, one study in older adults found tau pathology to mediate the correlation between LC signal and memory performance.³²

While progress has been made including several longitudinal studies of LC imaging in clinical or preclinical AD,^32,33,41 more work is needed to characterize the time course of LC degeneration, especially relative to the first presence of tau and Aβ, and its association with cognitive decline. Thus, the current study aims to examine LC signal loss longitudinally in a large sample spanning older adults with and without AD pathology. We quantify the LC NM–MRI signal using our validated automated LC segmentation method,⁴² and test the correlations between a summary measure of LC integrity and multiple cognitive measures across and within different groups of older adults stratified based on PET-imaging markers of Aβ and tau burden. The cognitive measures examined include global cognition, verbal memory, working memory, executive function, speed of processing, object recognition, and verbal fluency.

Methods

Participants

Study participants (n = 269), over 50 years of age from the community or outpatients at the McGill University Research Centre for Studies in Aging were enrolled in the Translational Biomarkers of Aging and Dementia (TRIAD) cohort,⁴³ McGill University, Canada. Participants had a detailed clinical assessment, including the Clinical Dementia Rating Scale (CDR) and Mini Mental State examination (MMSE; for clinical and demographic description of the sample, see Table 1). Cognitively unimpaired participants had no objective cognitive impairment and a CDR score of 0. Mild cognitive impairment (MCI) individuals had subjective and objective cognitive impairment, preserved activities of daily living, and a CDR score of 0.5. Patients with mild-to-moderate sporadic AD dementia had a CDR score between 0.5 and 2 and met the National Institute on Aging and the Alzheimer’s Association criteria for probable AD determined by a physician.⁴⁴ Participants were excluded if they had other inadequately treated conditions, active substance abuse, recent head trauma, major surgery, or MRI/PET safety contraindication. AD patients did not discontinue medications for this study. The Douglas Institute Research Ethics Board approved this study. Study procedures were carried out consistent with the Declaration of Helsinki principles and all participants provided written informed consent.

Table 1.

Participant characteristics.

Measure	Aβ− (n = 150)	Aβ+ (n = 119)	p Value
Age	70.9 ± 6.2	70.7 ± 7.1	0.78
Male, sex (%)	57 (38.0%)	44 (36.9%)	0.86
AD severity (CDR; 0/0.5/1/⩾2)	117/29/3/0	35/57/16/9	<0.001
Years of education	15.8 ± 3.5	15.1 ± 3.5	0.082
APOE ε4 carrier (%)	30 (20.0%)	54 (45.3%)	<0.001
Clinical diagnosis (CN/AD/MCI)	118/3/29	35/40/44	<0.001
LC signal (middle LC section contrast ratio, %)	23.2 ± 5.2	20.9 ± 4.8	<0.001
Tau burden, Braak stage 1 ROI ([¹⁸F]MK6240 SUVR)	0.93 ± 0.38	1.95 ± 0.83	<0.001
Tau burden, Braak stage 4 ROI ([¹⁸F]MK6240 SUVR)	1.09 ± 0.46	1.80 ± 1.00	<0.001
Amyloid β burden, neocortex ([¹⁸F]AZD4694 SUVR)	1.27 ± 0.10	2.29 ± 0.42	<0.001
Global cognition (MMSE)	29.08 ± 1.53	25.85 ± 5.18	<0.001
Logical memory (WAIS)	15.33 ± 4.84	8.37 ± 7.18	<0.001
Working memory (WAIS digit span)	16.32 ± 3.94	14.38 ± 4.73	<0.001
Speed of processing (TMT-A)	52.63 ± 29.87	40.76 ± 29.32	0.013
Executive functioning (TMT-B)	56.32 ± 28.55	37.74 ± 29.44	<0.001
Verbal fluency	39.64 ± 10.32	32.48 ± 13.40	<0.001
Object recognition (BORB)	30.43 ± 1.79	28.66 ± 3.76	<0.001
Mean months between assessment time points (range)	14.0 (5.3–27.9)	12.9 (6.0–25.9)	0.19
Sample size with (1/2/3/4) assessment time points	67/61/20/2	63/40/13/3	0.45
Mean assessments per participant	1.72 ± 0.74	1.64 ± 0.78	0.35

Mean ± standard deviation is displayed for continuous measures. p value are from t-tests for continuous variables and chi-square tests for categorical variables.

Cognitive measures

The Mini-Mental State Examination (MMSE) evaluates orientation, memory, attention, language, and visuospatial skills. We employ the total score reflecting performance on all tasks.⁴⁵ The Birmingham Object Recognition Battery (BORB) assesses object recognition using 32 items consisting of drawings of tools and animals. Half of the pictures depict a chimera of two animals (e.g. head of a donkey on a cow’s body). Participants must recognize whether a real or an unreal object is depicted.⁴⁶ The Trail Making Test (TMT) consists of part A (TMT-A) wherein participants connect numbers sequentially on a sheet of paper and part B where numbers and letters are connected in alternating sequence. Scoring considers speed and accuracy. The Digit Span Test is a component of the Wechsler Adult Intelligence Scale—Third Edition (WAIS-III) wherein participants repeat a sequence of digits in the forward and reverse order.⁴⁷ Logical memory is a subtest in the Wechsler Memory Scale-IV (WMS-IV) wherein participants listen to stories and recall as many details as possible. We used the immediate recall score (logical memory I).⁴⁸ The verbal fluency test requires participants to produce as many words as possible within a semantic category (category fluency condition) and starting with a given letter (letter fluency). We employ the combined score from both conditions.⁴⁹

MRI acquisition

All neuroimaging data were acquired at the Montreal Neurological Institute (MNI). Magnetic resonance (MR) images were acquired on a 3 T Prisma-Fit Siemens scanner. NM–MRI images were collected via a turbo spin echo sequence with the following parameters: repetition time (TR) = 600 ms; echo time (TE) = 10 ms; flip angle = 120°; turbo factor = 4; in-plane resolution = 0.6875 × 0.6875 mm²; partial brain coverage overlaying the pons and midbrain with field of view = 165 × 220; number of slices = 20; slice thickness = 1.8 mm; number of averages = 7; acquisition time = 8.45 min. Whole-brain, T1-weighted MR images (resolution = 1 mm, isotropic) were acquired using an MPRAGE sequence for preprocessing of the NM–MRI and PET data. Quality of MR images was visually inspected for artifacts immediately upon acquisition, and scans were repeated when necessary, time permitting.

NM–MRI processing

LC signal was measured using a semi-automated algorithm developed by our group (see previous papers for more complete description of the algorithm and performance of the method in terms of reliability and performance in the context of AD^42,50). This “Funnel-Tip” method performs an intensity-threshold-free cluster search within an overinclusive mask of the LC that has been transformed from standardized space to native space. It thereby estimates the signal from the LC at its hyperintense putative “core” on each cross-sectional (axial) slice through the structure. Similar to other algorithms, it calculates contrast-to-noise ratio (CNR) of the LC relative to a central pons reference region, applies an LC search mask on native space NM–MRI images⁵¹ and divides the LC into rostrocaudal sections⁵² (we calculated LC signal for five sections on the right and left side; see Figure 1(a) for LC segmentation in a representative participant).

Figure 1.

LC segmentation and measurement of LC signal in rostrocaudal sections: (a) LC segmentation for a representative subject using the Funnel-Tip method. A single axial slice shows the LC in cross-section (in red, four voxels per side on each axial slice covered by the overinclusive LC mask, only the most hyperintense of the four is retained), (b) schematic map of the LC showing weights used to derive the summary measure of LC integrity for each participant (LC pattern_CN). The colors reflect coefficients for each LC section derived from a logistic regression predicting CN status. High signal in the middle of the LC relative to its rostral and caudal ends is associated with greater odds of a participant being CN. These coefficients were then applied to the sectionwise LC values from each participant to calculate their LC pattern_CN value, and (c) longitudinal change in LC signal across rostrocaudal sections in sub-groups of participants at different stages of AD.

We used a multivariate approach, guided by work in other neuroimaging modalities,⁵³ to generate a single metric of LC integrity that incorporates information from all LC sections and reflects the extent to which an individual’s pattern of LC signal across sections is AD-like or healthy (CN-like). To make this metric, we first carried out a mixed-effects logistic regression analysis (fitglme with “binomial” distribution and “logit” link function in MATLAB) with clinical diagnosis of AD as the dependent variable (MCI were excluded) and LC signal from all 10 LC sections for all time points as fixed factors and subject as a random factor. The regression coefficients produced by this analysis for each LC section was saved (see Figure 1(b)). These same 10 regression coefficients were then input with each participant’s LC signal data from the corresponding LC section (using the mnrval MATLAB function) to return each subject’s predicted probability of CN status from a multinomial logistic regression model. Finally, the probabilities were converted to log of the odds ratio (as log of odds are more amenable to statistical analysis compared to probabilities) log of odds of CN status = log(probability(CN) / probability(AD)). All steps subsequent to calculation of coefficients applied to all participants including those with MCI. We refer to this log of the odds ratio as the LC pattern_CN, participants with positive values have greater odds of being CN than AD. Figure 1(b) shows that the LC pattern linked to CN status is high signal in middle and mid-caudal LC sections relative to signal on the rostral and caudal ends of the LC. Note that LC pattern_CN was highly correlated with a simpler LC integrity metric, LC signal averaged from bilateral middle LC sections (r = 0.77, the middle section was selected because it was the one that was most affected in AD in our prior work in this cohort^42,50).

PET acquisition and analysis

Participants had [¹⁸F]AZD4694 and [¹⁸F]MK6240 PET scans acquired to estimate brain levels of Aβ and tau, respectively, using a brain-dedicated Siemens high resolution research tomograph. PET measures were employed, not as outcomes of interest, but to stratify participants into groups based on AD severity. See previous studies for more detailed PET methods.^54,55

Briefly, [¹⁸F]MK6240 acquisition proceeded 90–110 min following intravenous bolus injection. Scans were reconstructed using an ordered-subsets expectation maximization (OSEM) algorithm on a four-dimensional volume with four frames (4 × 300 s). [¹⁸F]AZD4694 radioligand acquisition proceeded 40–70 min following intravenous bolus injection. Reconstruction used OSEM on a four-dimensional volume with three frames (3 × 600 s). For attenuation correction, PET acquisition was followed by a 6-min transmission scan conducted with a rotating 137Cs point source. Images were corrected for motion, dead time, radioactive decay as well as random and scattered coincidences and were meninges and skull stripped. PET images underwent linear registration to native T1 space (using ANTs) and non-linear transformation to ADNI reference space followed by smoothing using an 8 mm full-width half-maximum Gaussian kernel. To quantify radioligand binding, standardized uptake value ratios (SUVR) were computed using the inferior cerebellar GM as a reference region in the case of tau-PET [¹⁸F]MK6240 and whole cerebellar GM in the case of amyloid-PET [¹⁸F]AZD4694.

For the Aβ tracer, [¹⁸F]AZD4694, SUVR values were extracted from a composite set of regions including the precuneus, prefrontal, orbitofrontal, parietal, temporal, anterior, and posterior cingulate cortices.⁵⁶ Aβ positivity (Aβ status) was defined, similar to prior work,⁵⁵ as SUVR value >1.55 at baseline or at a subsequent time point if [¹⁸F]AZD4694 was not available at baseline. SUVR scores from the tau tracer [¹⁸F]MK6240 were extracted from predefined ROIs to determine Braak stage, consistent with prior work.⁵⁷ Each ROI is linked to a specific Braak stage and each participant’s Braak stage was defined based on the most advanced stage ROI where SUVR was above that ROI’s critical threshold (for further detail about threshold determination, see prior work⁵⁷). Discordant observations (where regional tau burden did not follow the anatomical progression proposed by Braak) were excluded from analyses of Braak stage, see Table S1 for characteristics of participants who were excluded from the Braak stage subgroups at baseline. Mean time interval across participants between [¹⁸F]AZD4694 and [¹⁸F]MK6240 scans was 40 days; mean interval between MRI and [¹⁸F]MK6240 scans was 38 days and between MRI and [¹⁸F]AZD4694 scans was 46 days.

Statistical analysis

Statistical analyses employed linear mixed effects models and robust linear regression implemented in MATLAB software with random intercepts but not random slopes. Our initial analysis examined the impact of time on the LC signal. This linear mixed-effects model employed an LC pattern_CN as dependent variable, a random subject factor, and the fixed factors were: time since baseline (time), Aβ status, time × Aβ status, age at baseline, sex, and scan date pre/post covid-19 pandemic onset. The pre/post covid-19 dummy covariate was included in analyses examining time effects on the LC because the pandemic occurred between the baseline and follow up scans of many participants and there is evidence that the LC may be vulnerable to covid-19 infection.^58–62 This analysis was repeated in different subgroups of participants examining the impact of time on LC signal with the same model except for exclusion of the Aβ status and time × Aβ status fixed factors.

The next set of analyses examined the impact of LC signal on cognition. These linear mixed-effects models employed a cognitive measure as dependent variable a random subject factor, and the following fixed factors: LC pattern_CN, age at study baseline, years of education, sex, and time. These were repeated for different cognitive measures and in different subgroups of participants.

For purposes of visualization, we provide statistical maps displaying the time effect (Figure 1(c)) and correlations to cognition (Figure S1) for all LC sections. These were generated by running the above-described linear-mixed effects regression models for all 10 LC sections.

Results

Longitudinal change in LC signal

To assess LC change over time, n = 269 study participants were included in a linear mixed effects model. We did not see evidence that the LC pattern_CN was differentially changing over time between Aβ-positive and Aβ-negative individuals (time × Aβ status interaction t₄₇₄ = −0.47). Therefore, we reasoned that further division into early and later phase AD groups may be needed to reveal AD-associated LC degeneration over time. We divided cases into a healthy aging group with no cognitive impairment and no AD pathophysiology (group (CN, Aβ-negative, and at Braak stage 0, n = 93, n = 41 with multiple time points), an early AD group group (Aβ-positive and Braak stage <4, n = 67, n = 27 with multiple time points), and a later AD group (Aβ-positive and Braak stage 4–6, n = 59, n = 21 with multiple time points). We did not measure significant change over time in the healthy aging group (ΔLC pattern_CN/year = −0.023, CI = −0.24 to 0.19, t₁₃₉ = −0.21, p = 0.83, linear mixed effects model controlling for age at baseline, sex, and scan date prior to covid-19 pandemic), nor in the later AD group (ΔLC pattern_CN/year = 0.13, CI = −0.096 to 0.35, t₈₀ = 1.14, p = 0.26). It was the early-stage AD group who showed a marked decrease in LC integrity over time (ΔLC pattern_CN/year = −0.43, CI = −0.66 to −0.20, t₉₅ = −3.72, p = 0.0003). Note that examining a simpler metric of LC integrity, the average signal from the bilateral middle LC, the results were similar (t₁₃₉ = −1.27, p = 0.21 in healthy aging; t₉₅ = −3.74, p = 0.0003 in early AD; t₈₀ = −1.23, p = 0.22 in later AD). For detailed visualization of how the LC signal changed over time in all rostro-caudal sections, see Figure 1(c) and for plots of LC integrity over time for each participant in the three groups, see Figure 2.

Figure 2.

Longitudinal change in LC integrity over time at different stages of AD. Each line plots LCpattern_CN at all assessment times for each participant. LC integrity significantly decreased over the 2–4 years of follow-up for participants at the early stages of AD (AB+, Braak <4, middle plot).

Association of LC integrity to cognition

Next, we considered the association between LC integrity and multiple domains of cognition in healthy aging and AD individuals using data from all available time points in linear mixed effects models with additional fixed factors: time, years of education, age at baseline, and sex. When merging Aβ-positive individuals from Braak stage 1–6, all cognitive measures correlated positively to LC pattern_CN, the strongest association was with the novel object recognition test (BORB: T₁₂₄ = 4.08, p = 0.00008; Table 2). In the healthy aging group, only working memory (WAIS digit span) showed a significant correlation to LC pattern_CN (T₁₂₈ = 2.62, p = 0.010; Table 2). To minimize the risk of correlation strength being inflated by examining measures across such a broad range of AD severity, and in line with evidence from the prior analyses that LC signal may be impacted differently across the AD spectrum, we performed subsequent analysis dividing the AD group into early (Aβ-positive, Braak stage <4) and later (Aβ-positive, Braak stage 4–6) groups. In the early AD group, LC pattern_CN correlated positively to general cognitive impairment (MMSE score, t₉₂ = 2.34, p = 0.021) and executive function (TMT-B test, t₄₇ = 2.92, p = 0.005; Table 2). In the later AD group, LC pattern_CN correlated positively to several measures including object recognition (BORB test, t₆₃ = 3.42, p = 0.001), logical memory (t₆₃ = 2.38, p = 0.020), and verbal fluency (t₆₅ = 2.33, p = 0.023; Table 2). The findings from examining cognitive domains in AD subgroups that survived FDR correction were the relationship of LC pattern_CN to TMT-B in early AD and its relationship to object recognition, logical memory, and verbal fluency in later AD. See Figure 3 for scatterplots of these relationships and Figure S1 for maps visualizing the relationships across LC sections. These analyses were repeated using a simpler LC integrity metric, the bilateral average of the middle LC section signal, and results were very similar compared to the LC pattern_CN metric (see Table S2).

Table 2.

Relation of LC pattern_CN to cognitive measures in study subgroups.

Cognitive measure	CN and Braak 0 and AB-	AB+, tau +	Braak <4 and AB+	Braak 4–6 and AB+
Global cognition	T₁₃₆ = −0.11	T₁₄₀ = 2.12*	T₉₂ = 2.34*	T₇₇ = 1.67
Logical memory	T₁₂₉ = 1.45	T₁₂₄ = 2.67**	T₈₄ = −0.10	T₆₃ = 2.38*
Working memory	T₁₂₈ = 2.62**	T₁₂₄ = 2.72**	T₈₃ = 0.28	T₆₄ = 2.11*
Speed of processing	T₇₀ = 1.41	T₅₅ = 2.65*	T₄₇ = 1.97	T₂₁ = 1.09
Executive functioning	T₇₀ = 1.18	T₅₂ = 3.47**	T₄₇ = 2.92**	T₁₈ = 0.00
Verbal fluency	T₁₂₉ = 1.45	T₁₂₆ = 2.89**	T₈₄ = 1.07	T₆₅ = 2.33*
Object recognition	T₁₂₈ = −0.40	T₁₂₄ = 4.08***	T₈₄ = 1.66	T₆₃ = 3.42**

p < 0.05. **p < 0.01. ***p < 0.001.

Figure 3.

Correlation between LC integrity and selected cognitive measures. Baseline assessments are shown as blue circles and follow-up assessments as orange circles. The correlation to executive function is shown for participants at the early stages of AD (Braak <4 and for amyloid β positive, AB+, at left) and the other correlations are shown for later AD participants (Braak >3 and AB+).

Given the strong relationship we observed in AD between object recognition and LC integrity, we further examined subscale scores of this test and found that the relationships to LC integrity were apparent when attempting to recognize novel stimuli (imaginary animals comprising the “unreal subscale,” t₁₂₄ = 4.01 in the full AD sample and t₆₃ = 3.59 the later AD group; Figure 3) but were not significant when recognizing familiar stimuli (common animals on the “real subscale,” t = 1.96 and 1.18, respectively).

Discussion

This study investigated longitudinal change in LC integrity and its association to cognition in a large and well characterized sample of older adults across the AD spectrum. We found that the rate of loss in LC integrity was steepest in the early stages of AD (prior to Braak stage 4), whereas a significant change could not be detected in later stages of AD or in healthy aging. Next, we examined the relationship of LC integrity to cognition across different domains. We found that, when looking at the full spectrum of AD severity, LC integrity was related to performance on all cognitive measures. A more stringent test to reduce spurious correlation due to illness progression and constrain analyses to either early AD (Braak stage <4) or later AD (Braak stage >3) found that many cognitive measures remained significantly correlated to LC integrity. While most of these cognitive measures have been previously implicated in the LC’s role in AD (e.g. logical memory^32,39,63), the one that showed the strongest relationship, object recognition, has been less well studied in this context and could reflect an important role of the LC–NE system in impaired novelty detection in AD. Finally, our study was able to investigate healthy aging according to a more rigorous definition than some prior reports that lacked Aβ and tau imaging. We found that, except for working memory, the LC metric was not significantly related to cognition in this group, consistent with evidence that LC degeneration is minimal in healthy aging.²⁶

This study generated a single measure of LC integrity across all rostrocaudal LC sections, LC pattern_CN—the log of the odds that an individual is cognitively normal or has AD based on their LC signal. This multivariate approach has some advantages over examining an average of the whole LC or individual LC sections without considering information provided by other sections. For example, a person with low LC signal in the middle of the LC relative to its ends will have a worse LC pattern_CN compared to another person with similarly low LC signal but throughout the LC. We had reasoned that diverse factors could influence the LC signal and a multivariate LC metric would assist in capturing LC signal loss over time as opposed to other biological or imaging artifacts might alter LC signal with a different pattern across LC sections compared to AD-linked degeneration. Indeed, this strategy revealed that the strongest LC degeneration occurred in the early phases of AD, from Braak stage 1–3, and it is notable that when examining LC signal change over time across all LC sections at this phase (Figure 1(c)), as expected, the pattern appears as the inverse of the pattern linked to CN status cross-sectionally (LC pattern_CN, see Figure 1(b)): loss in the middle of the LC but trending towards increase on the ends of the LC.

The observation of decreasing LC signal in early AD is consistent with prior work examining the early or at-risk stage of AD. We are aware of two prior imaging studies on this question, both of which found significant LC signal decrease, either in people at genetic risk for AD³² or in a mixed group of AD and MCI.³³ Postmortem work has likewise shown the sharpest drop in LC volume between Braak stage 0 and 1 in middle LC.²⁶ On the other hand, another study on a large sample of community older adults did not show a significant change in LC signal over time.³⁹ This lack of longitudinal change in LC signal in healthy aging (or in the later AD group) could be accounted for by several factors. Small changes in LC signal may not rise above the level of test-retest noise, possibly requiring a longer follow-up than the 2–4-year interval to observe significant change. Furthermore, we cannot directly measure LC degeneration with MRI but only LC signal, which is susceptible to other influences that could plausibly increase the signal (e.g. changes in neuromelanin, iron, cell size or macromolecule content) and counteract a degeneration-linked decrease. Indeed, we observed large increases in signal over time for some participants (Figure 2). This could obscure LC degeneration and thus we cannot conclude whether our inability to observe signal decrease over time in the later AD group was due to a flattening rate of degeneration or due to these other influences assuming that they could fluctuate more widely at later illness stages. Indeed, our multivariate LC metric may have been better suited to observe change in the early, compared to later AD stages and it is possible that the pattern of LC signal loss throughout the structure may vary depending on the stage of AD. Nevertheless, examining both the summary metric and the more granular LC metrics (Figure 1(c)) it appears that LC subregions show a steeper rate of signal loss in early AD than later AD. This still leaves open the possibility that more total LC loss occurs in later AD should the duration of these late AD stages be much longer than the early stages (lower rate but for a longer duration could equal more loss). Other important caveats when considering early versus later AD include the fact that the later AD group was the smallest in the study (only n = 21 with follow up data) and may have been underpowered to detect signal loss over time and finally that participants in our study did not have severe AD symptoms and it is possible that heavy LC loss may occur in the latest stages of illness.

We found LC integrity to be positively associated with cognitive performance across several measures, consistent with prior imaging studies. To examine more directly the association of LC signal with cognition and minimize spurious correlation from sampling across a wide range of AD severity, we looked at early and later AD subgroups and still found many significant correlations. In the early AD subgroup, the only measure to withstand multiple-comparisons correction was the TMT-B test. This finding is consistent with a large body of evidence of the LC-NE system’s role in executive function⁶³ and merits replication in a larger sample (this test was available on fewer participants than most tests in our sample) and using a paradigm capable of investigating specific aspects of executive function (e.g. cognitive flexibility, inhibitory control). In the later AD group, we found several cognitive measures to positively correlate to LC integrity including logical memory, a common measure of interest in prior LC imaging studies, most of which also found LC signal loss linked to deficits in logical memory in AD or AD-risk groups.^31,37,63 Our observation of a correlation between LC signal and verbal fluency is consistent with a prior report in older adults with and without MCI.³⁶ It is notable that both the logical memory and verbal fluency effects appeared to be most pronounced in the mid-caudal LC (Figure S1), a topography mirroring a prior report.³⁶

It is notable that the cognitive measure that related most strongly to LC integrity, object recognition, has not been investigated in prior LC imaging studies to our knowledge. This measure correlated only at the later AD stage perhaps due to ceiling effects in those with lesser impairment as near-perfect scores would be expected for most healthy adults on a test of this nature. We argue that this measure can be considered a novelty detection test by asking participants to recognize pictures of animals that do not exist in nature. There is a large literature on the role of the LC-NE system on novelty detection and on the presence of novelty detection deficits in AD.^64–69 Perhaps this is a particularly sensitive test to demonstrate the impact of LC damage on behavior in AD, and could be considered as a simple screening test to track LC integrity in AD. Furthermore, more detailed study of the LC on novelty processing could better reveal mechanisms of norepinephrine system dysfunction in AD and how they may color the expression of behavioral impairment.

We found that in healthy aging, with the exception of working memory, our LC metric was not significantly related to cognition, consistent with the absence of significant LC signal loss over time in this group and prior evidence that LC degeneration is minimal in healthy aging based on postmortem and imaging data.^26,31,67,68 Thus, the factors contributing to LC signal variability in the absence of LC degeneration (e.g. neuromelanin and iron content, cell size or cell density) may have little consequence on cognitive performance in healthy aging. Several prior studies^31,32,68,69 similarly found no significant association of LC signal to cognitive performance in healthy aging, whereas other studies did find LC signal to be correlated with memory,^29,39,70 global cognition,³⁸ and executive function.⁴⁰ These associations may in some cases be due to inadvertent inclusion of cases with prodromal AD as many studies did not screen for brain AD markers. This ambiguity encouraged us to be very strict in our definition of healthy aging (cognitively normal and no Aβ or tau, even in the Braak 0 ROI). Nevertheless, we do not mean to conclude that the LC is immune to harms in healthy aging just because we did not observe change in our measures over the time of follow-up. Aside from the strong evidence of early tau accumulation in LC, some imaging and postmortem studies have observed LC damage in this context⁷¹ and the susceptibility of LC neurons to degeneration in Braak stage 0 remains a subject of debate. Regarding correlation to the correlation of LC integrity to working memory, this finding is intriguing, but contrasts with some prior work in healthy aging,³⁹ and merits replication.

Our study has many strengths as the largest longitudinal LC imaging study we are aware of in AD and one that spans the full spectrum of AD from Braak stage 0–6 and includes PET imaging of tau and Aβ. Nevertheless, not all subjects had longitudinal data and, after division into subgroups, sample sizes were still modest. Another limitation is a risk of circularity considering that our LC metric, LC pattern_CN, was defined and applied in the same sample. We feel this risk is minimal in our case since this metric was defined by a group comparison (CN vs AD) and very few analyses in the study were performed across groups and the most important findings were observed in analyses confined to participants largely drawn from one or the other of these groups. Future work could apply the same sectionwise weights used in this study to calculate LC pattern_CN in an out-of-sample study to determine whether this metric would allow better diagnostic segregation compared to a simpler metric (e.g. whole LC or middle LC signal). Furthermore, since this metric was defined to segregate CN from AD, it may not allow optimal distinction within AD, that is, patterns of LC signal that differentiate early from later AD stages. Thus, future work is needed to better determine the utility of this multivariate metric and it is reassuring that in the current report the findings using the multivariate metric of LC integrity were closely aligned to findings using a simpler and more standard metric of LC contrast in the middle section. Finally, our study found the largest signal loss occurring in the middle LC section, in contrast to some other reports where the largest effect was in mid-rostral LC.^29,33,39 This discrepancy could be due to methodological differences between studies, for instance in the anatomical definition of the LC or MRI acquisition, or differences in sample characteristics.

In conclusion, our study reveals that LC signal loss is detectable within subjects over a 2–4-year span in early stages of AD. Furthermore, we confirm prior reports of positive correlations of LC signal to multiple cognitive domains in AD but show that LC signal variation in healthy aging may have little impact on cognition. Finally, we propose that tests of novel object recognition may be particularly sensitive to LC damage in AD.

Supplemental Material

sj-docx-1-jcb-10.1177_0271678X261460160 – Supplemental material for A longitudinal investigation of locus coeruleus norepinephrine system integrity and cognitive performance across the Alzheimer’s spectrum

Supplemental material, sj-docx-1-jcb-10.1177_0271678X261460160 for A longitudinal investigation of locus coeruleus norepinephrine system integrity and cognitive performance across the Alzheimer’s spectrum by Ahmad Sibahi, Seyyed Ali Hosseini, Nesrine Rahmouni, Joseph Therriault, Rami Hamati, Reihaneh Forouhandehpour, Firoza Lussier, Mira Chamoun, Christine Tardif, Pedro Rosa-Neto and Clifford M Cassidy in Journal of Cerebral Blood Flow & Metabolism

Footnotes

Acknowledgements

The authors thank all participants of the study and staff of the McGill Center for studies in Aging and at the University of Ottawa Institute of Mental Health Research.

Author contributions

CMC, PR-N, and AS made substantial contributions to the conception and design of the work, to the acquisition, analysis, or interpretation of data for the work; and to drafting of the work and revising it critically for important intellectual content. MC, NR, and FL contributed to collection of neuroimaging and/or clinical data. CT contributed to implementation and analysis of neuroimaging measures. JT, SAH, RH, and RF contributed to data management, data processing and analysis. All authors contributed to writing and editing the manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is supported by the Natural Sciences and Engineering Research Council of Canada (NSERC; RGPIN-2020-06127, PI: CMC), the Weston Brain Institute, Canadian Institutes of Health Research (CIHR; MOP-11-51-31; RFN 152985, 159815, 162303), Canadian Consortium of Neurodegeneration and Aging (CCNA; MOP-11-51-31-team 1), the Alzheimer’s Association (NIRG-12-92090, NIRP-12-259245), Brain Canada Foundation (CFI Project 34874; 33397), the Fonds de Recherche du Québec—Santé (FRQS; Chercheur Boursier, 2020-VICO-279314; 2024-VICO-356138). PR-N is a member of the CIHR-CCNA Canadian Consortium of Neurodegeneration in Aging. Colin J Adair Charitable Foundation.

Declaration of conflicting interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: The authors report no competing financial interest in relation to the study design, results, or discussion. CMC and PR-N are inventors on a pending patent using the analysis method described here, licensed to Terran Biosciences, but have received no royalties.

ORCID iD

Ahmad Sibahi

Supplemental material

Supplemental material for this article is available online.

References

Sara

SJ.

The locus coeruleus and noradrenergic modulation of cognition. Nat Rev Neurosci 2009; 10: 211–223.

Aston-Jones

Cohen

JD.

An integrative theory of locus coeruleus–norepinephrine function: adaptive gain and optimal performance. Annu Rev Neurosci 2005; 28: 403–450.

Poe

Foote

Eschenko

, et al. Locus coeruleus: a new look at the blue spot. Nat Rev Neurosci 2020; 21: 644–659.

Sara

Bouret

Orienting and reorienting: the locus coeruleus mediates cognition through arousal. Neuron 2012; 76: 130–141.

Mather

Harley

CW.

The locus coeruleus: essential for maintaining cognitive function and the aging brain. Trends Cogn Sci; 20. Epub ahead of print 2016. DOI: 10.1016/j.tics.2016.01.001.

Clewett

Lee

Greening

, et al. Neuromelanin marks the spot: identifying a locus coeruleus biomarker of cognitive reserve in healthy aging. Neurobiol Aging; 37. Epub ahead of print 2016. DOI: 10.1016/j.neurobiolaging.2015.09.019.

Marien

Colpaert

Rosenquist

AC.

Noradrenergic mechanisms in neurodegenerative diseases: a theory. Brain Res Rev 2004; 45: 38–78.

Polak

Kalinin

Feinstein

DL.

Locus coeruleus damage and noradrenaline reductions in multiple sclerosis and experimental autoimmune encephalomyelitis. Brain 2011; 134: 665–677.

Ohtsuka

Sasaki

Konno

, et al. Changes in substantia nigra and locus coeruleus in patients with early-stage Parkinson’s disease using neuromelanin-sensitive MR imaging. Neurosci Lett 2013; 541: 93–98.

10.

Langley

Huddleston

Liu

, et al. Reproducibility of locus coeruleus and substantia nigra imaging with neuromelanin sensitive MRI. Magn Reson Mater Phys Biol Med 2017; 30: 121–125.

11.

Reimão

Pita Lobo

Neutel

, et al. Substantia nigra neuromelanin magnetic resonance imaging in de novo Parkinson’s disease patients. Eur J Neurol; 22. Epub ahead of print 2015. DOI: 10.1111/ene.12613.

12.

Bari

Chokshi

Schmidt

Locus coeruleus–norepinephrine: basic functions and insights into Parkinson’s disease. Neural Regen Res 2020; 15: 1006–1013.

13.

Ehrenberg

Nguy

Theofilas

, et al. Quantifying the accretion of hyperphosphorylated tau in the locus coeruleus and dorsal raphe nucleus: the pathological building blocks of early Alzheimer’s disease. Neuropathol Appl Neurobiol 2017; 43: 393–408.

14.

Byun

Jung

, et al. Locus coeruleus tau is linked to successive cortical tau accumulation. Alzheimers Dement. Epub ahead of print 1 February 2024. DOI: 10.1002/alz.14426.

15.

Ghosh

Torraville

Mukherjee

, et al. An experimental model of Braak’s pretangle proposal for the origin of Alzheimer’s disease: the role of locus coeruleus in early symptom development. Alzheimers Res Ther; 11. Epub ahead of print 3 July 2019. DOI: 10.1186/s13195-019-0511-2.

16.

Chan-Palay

Asan

Alterations in catecholamine neurons of the locus coeruleus in senile dementia of the Alzheimer type and in Parkinson’s disease with and without dementia and depression. J Comp Neurol 1989; 287: 373–392.

17.

Bekdash

RA.

The cholinergic system, the adrenergic system and the neuropathology of Alzheimer’s disease. Int J Mol Sci 2021; 22: 1–18.

18.

Grudzien

Shaw

Weintraub

, et al. Locus coeruleus neurofibrillary degeneration in aging, mild cognitive impairment and early Alzheimer’s disease. Neurobiol Aging 2007; 28: 327–335.

19.

McMillan

White

Franklin

, et al. Differential response of the central noradrenergic nervous system to the loss of locus coeruleus neurons in Parkinson’s disease and Alzheimer’s disease. Brain Res 2011; 1373: 240–252.

20.

Arendt

Brückner

Morawski

, et al. Early neurone loss in Alzheimer’s disease: cortical or subcortical?. Acta Neuropathol Commun 2015; 3: 10.

21.

Theofilas

Dunlop

Heinsen

, et al. Turning on the light within: subcortical nuclei of the isodentritic core and their role in Alzheimer’s disease pathogenesis. J Alzheimers Dis 2015; 46: 17–34.

22.

James

Wilson

Boyle

, et al. P3–192: life-span cognitive activity, neuropathologic burden and cognitive aging. Alzheimers Dement; 9. Epub ahead of print July 2013. DOI: 10.1016/j.jalz.2013.05.1264.

23.

Ghasemi

Mehranfard

Neuroprotective actions of norepinephrine in neurological diseases. Pflugers Arch 2024; 476: 1703–1725.

24.

Chen

Huddleston

Langley

, et al. Simultaneous imaging of locus coeruleus and substantia nigra with a quantitative neuromelanin MRI approach. Magn Reson Imaging; 32. Epub ahead of print 2014. DOI: 10.1016/j.mri.2014.07.003.

25.

Cassidy

Zucca

Girgis

, et al. Neuromelanin-sensitive MRI as a noninvasive proxy measure of dopamine function in the human brain. Proc Natl Acad Sci U S A; 116. Epub ahead of print 2019. DOI: 10.1073/pnas.1807983116.

26.

Theofilas

Ehrenberg

Dunlop

, et al. Locus coeruleus volume and cell population changes during Alzheimer’s disease progression: a stereological study in human postmortem brains with potential implication for early-stage biomarker discovery. Alzheimers Dement; 13. Epub ahead of print 2017. DOI: 10.1016/j.jalz.2016.06.2362.

27.

Weinshenker

Long road to ruin: noradrenergic dysfunction in neurodegenerative disease. Trends Neurosci; 41. Epub ahead of print 2018. DOI: 10.1016/j.tins.2018.01.010.

28.

Castellanos

Fernández-Seara

Lorenzo-Betancor

, et al. Automated neuromelanin imaging as a diagnostic biomarker for Parkinson’s disease. Mov Disord 2015; 30: 945–952.

29.

Dahl

Mather

Düzel

, et al. Rostral locus coeruleus integrity is associated with better memory performance in older adults. Nat Hum Behav; 3. Epub ahead of print 2019. DOI: 10.1038/s41562-019-0715-2.

30.

Dünnwald

Betts

Sciarra

, et al. Automated segmentation of the locus coeruleus from neuromelanin-sensitive 3T MRI using deep convolutional neural networks. In: Informatik aktuell, 2020. DOI: 10.1007/s11548-021-02528-5.

31.

Jacobs

HIL

Becker

Kwong

, et al. In vivo and neuropathology data support locus coeruleus integrity as indicator of Alzheimer’s disease pathology and cognitive decline. Sci Transl Med; 13. Epub ahead of print 22 September 2021. DOI: 10.1126/scitranslmed.abj2511.

32.

Jacobs

HIL

Becker

Kwong

, et al. Waning locus coeruleus integrity precedes cortical tau accrual in preclinical autosomal dominant Alzheimer’s disease. Alzheimers Dement 2023; 19: 169–180.

33.

Galgani

Lombardo

Frijia

, et al. The degeneration of locus coeruleus occurring during Alzheimer’s disease clinical progression: a neuroimaging follow-up investigation. Brain Struct Funct 2024; 229: 1317–1325.

34.

Galgani

Giorgi

FS.

Exploring the role of locus coeruleus in Alzheimer’s disease: a comprehensive update on MRI studies and implications. Curr Neurol Neurosci Rep 2023; 23: 925–936.

35.

Galgani

Lombardo

Martini

, et al. Magnetic resonance imaging locus coeruleus abnormality in amnestic mild cognitive impairment is associated with future progression to dementia. Eur J Neurol 2023; 30: 32–46.

36.

Elman

Puckett

Beck

, et al. MRI-assessed locus coeruleus integrity is heritable and associated with cognition, Alzheimer’s risk, and sleep–wake disturbance. Alzheimers Dement 2021; 17: 1017–1025.

37.

Wolters

Heijmans

Priovoulos

, et al. Neuromelanin related ultra-high field signal intensity of the locus coeruleus differs between Parkinson’s disease and controls. Neuroimage Clin; 39. Epub ahead of print 1 January 2023. DOI: 10.1016/j.nicl.2023.103479.

38.

Liu

Kievit

Tsvetanov

, et al. Noradrenergic-dependent functions are associated with age-related locus coeruleus signal intensity differences. Nat Commun; 11. Epub ahead of print 1 December 2020. DOI: 10.1038/s41467-020-15410-w.

39.

Dahl

Bachman

Dutt

, et al. The integrity of dopaminergic and noradrenergic brain regions is associated with different aspects of late-life memory performance. Nat Aging 2023; 3: 1128–1143.

40.

Giorgi

Lombardo

Galgani

, et al. Locus coeruleus magnetic resonance imaging in cognitively intact elderly subjects. Brain Imaging Behav 2022; 16: 1077–1087.

41.

Bueichekú

Diez

Kim

, et al. Spatiotemporal patterns of locus coeruleus integrity predict cortical tau and cognition. Nat Aging 2024; 4: 625–637.

42.

Sibahi

Gandhi

Al-Haddad

, et al. Characterization of an automated method to segment the human locus coeruleus. Hum Brain Mapp 2023; 44: 3913–3925.

43.

Therriault

Benedet

Pascoal

, et al. Association of apolipoprotein E ε4 with medial temporal tau independent of amyloid-β supplemental content. JAMA Neurol 2020; 77: 470–479.

44.

McKhann

Knopman

Chertkow

, et al. The diagnosis of dementia due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement; 7. Epub ahead of print 2011. DOI: 10.1016/j.jalz.2011.03.005.

45.

Shigemori

Ohgi

Okuyama

, et al. The factorial structure of the mini mental state examination (MMSE) in Japanese dementia patients. BMC Geriatr 2010; 10: 36.

46.

St-Hilaire

Blackburn

Wilson

, et al. Object decision test (BORB): normative data for the adult Quebec population and performance in Alzheimer’s disease and the semantic variant of primary progressive aphasia. Aging Neuropsychol Cogn 2018; 25: 427–442.

47.

Lichtenberger

Kaufman

AS.

Essentials of WAIS-IV assessment. Hoboken, NY: John Wiley & Sons, 2012.

48.

Ahn

Joung

, et al. Normative data for the logical memory subtest of the Wechsler Memory Scale-IV in middle-aged and elderly Korean people. Psychiatry Investig 2019; 16: 793–799.

49.

Shao

Janse

Visser

, et al. What do verbal fluency tasks measure? Predictors of verbal fluency performance in older adults. Front Psychol; 5. Epub ahead of print 2014. DOI: 10.3389/fpsyg.2014.00772.

50.

Cassidy

Therriault

Pascoal

, et al. Association of locus coeruleus integrity with Braak stage and neuropsychiatric symptom severity in Alzheimer’s disease. Neuropsychopharmacology 2022; 47: 1128–1136.

51.

García-Lorenzo

Longo-Dos Santos

Ewenczyk

, et al. The coeruleus/subcoeruleus complex in rapid eye movement sleep behaviour disorders in Parkinson’s disease. Brain 2013; 136: 2120–2129.

52.

Jacobs

HIL

Priovoulos

Poser

, et al. Dynamic behavior of the locus coeruleus during arousal-related memory processing in a multi-modal 7T fMRI paradigm. Elife; 9. Epub ahead of print 2020. DOI: 10.7554/eLife.52059.

53.

Horga

Cassidy

, et al. Dopamine-related disruption of functional topography of striatal connections in unmedicated patients with schizophrenia. JAMA Psychiatry 2016; 73: 862–870.

54.

Lussier

Pascoal

Chamoun

, et al. Mild behavioral impairment is associated with β-amyloid but not tau or neurodegeneration in cognitively intact elderly individuals. Alzheimers Dement; 16. Epub ahead of print 2020. DOI: 10.1002/alz.12007.

55.

Therriault

Benedet

Pascoal

, et al. Determining amyloid-β positivity using ¹⁸F-AZD4694 PET imaging. J Nucl Med 2021; 62: 247–252.

56.

Jack

Wiste

Weigand

, et al. Defining imaging biomarker cut points for brain aging and Alzheimer’s disease. Alzheimers Dement; 13. Epub ahead of print 2017. DOI: 10.1016/j.jalz.2016.08.005.

57.

Pascoal

Therriault

Benedet

, et al. ¹⁸F-MK-6240 PET for early and late detection of neurofibrillary tangles. Brain; 143. Epub ahead of print 2020. DOI: 10.1093/brain/awaa180.

58.

Hugon

Queneau

Sanchez Ortiz

, et al. Cognitive decline and brainstem hypometabolism in long COVID: a case series. Brain Behav; 12. Epub ahead of print 1 April 2022. DOI: 10.1002/brb3.2513.

59.

Rogers

Chesney

Oliver

, et al. Psychiatric and neuropsychiatric presentations associated with severe coronavirus infections: a systematic review and meta-analysis with comparison to the COVID-19 pandemic. Lancet Psychiatry 2020; 7: 611–627.

60.

Emmi

Rizzo

Barzon

, et al. Detection of SARS-CoV-2 viral proteins and genomic sequences in human brainstem nuclei. NPJ Parkinsons Dis; 9. Epub ahead of print 1 December 2023. DOI: 10.1038/s41531-023-00467-3.

61.

Martin

Rupprecht

Schrenk

, et al. A hypoarousal model of neurological post-COVID syndrome: the relation between mental fatigue, the level of central nervous activation and cognitive processing speed. J Neurol 2023; 270: 4647–4660.

62.

Jacobs

HIL

Ibrahim

Vahidy

, et al. Lower locus coeruleus integrity in older COVID-19 survivors: initial findings from an international 7T MRI consortium. Alzheimers Dement; 19. Epub ahead of print December 2023. DOI: 10.1002/alz.080224.

63.

Olivieri

Lagarde

Lehericy

, et al. Early alteration of the locus coeruleus in phenotypic variants of Alzheimer’s disease. Ann Clin Transl Neurol 2019; 6: 1345–1351.

64.

Schneider

Prokopiou

Papp

, et al. Atrophy links lower novelty-related locus coeruleus connectivity to cognitive decline in preclinical AD. Alzheimers Dement 2024; 20: 3958–3971.

65.

Bastin

Delhaye

Moulin

, et al. Novelty processing and memory impairment in Alzheimer’s disease: a review. Neurosci Biobehav Rev 2019; 100: 237–249.

66.

Omoluabi

Torraville

Maziar

, et al. Novelty-like activation of locus coeruleus protects against deleterious human pretangle tau effects while stress-inducing activation worsens its effects. Alzheimers Dement; 7. Epub ahead of print 2021. DOI: 10.1002/trc2.12231.

67.

Lustberg

Tillage

Bai

, et al. Noradrenergic circuits in the forebrain control affective responses to novelty. Psychopharmacology 2020; 237: 3337–3355.

68.

Prokopiou

Engels-Domínguez

Papp

, et al. Lower novelty-related locus coeruleus function is associated with Aβ-related cognitive decline in clinically healthy individuals. Nat Commun; 13. Epub ahead of print 1 December 2022. DOI: 10.1038/s41467-022-28986-2.

69.

Prokopiou

Engels-Domínguez

Schultz

, et al. Association of novelty-related locus coeruleus function with entorhinal tau deposition and memory decline in preclinical Alzheimer disease. Neurology 2023; 101: E1206–E1217.

70.

Al Haddad

Chamoun

Tardif

, et al. Normative values of neuromelanin-sensitive MRI signal in older adults obtained using a turbo spin echo sequence. J Magn Reson Imaging 2023; 58: 294–300.

71.

Beardmore

Hou

Darekar

, et al. The locus coeruleus in aging and Alzheimer’s disease: a postmortem and brain imaging review. J Alzheimers Dis 2021; 83: 5–22.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.07 MB