University of Southern California Dissertations and Theses

Age differences in diffusivity in the locus coeruleus and its ascending noradrenergic tract

(USC Thesis Other)

Age differences in diffusivity in the locus coeruleus and its ascending noradrenergic tract

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content RUNNING HEAD: AGING AND LC DIFFUSIVITY

AGE DIFFERENCES IN DIFFUSIVITY IN THE LOCUS COERULEUS AND ITS
ASCENDING NORADRENERGIC TRACT
by
Shai Porat

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(NEUROSCIENCE)
August 2022

RUNNING HEAD: AGING AND LC DIFFUSIVITY
ii
Dedication

To my mother who has always been fascinated by diffusion of water molecules in the
locus coeruleus and its ascending noradrenergic bundle. I’m just kidding, she doesn’t understand
what I’ve been doing for the last five years but supports me anyways.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
iii
Table of Contents
Dedication ...................................................................................................................................................................... ii
List of Tables .................................................................................................................................................................. v
List of Figures .............................................................................................................................................................. vii
Abstract ....................................................................................................................................................................... viii
Chapter 1: Age differences in locus coeruleus diffusivity and its ascending noradrenergic bundle .............................. 1
1. Introduction ...................................................................................................................................................... 1
2. Methods............................................................................................................................................................ 2
2.1. DWI processing ...................................................................................................................................... 8
2.2. Statistical analyses ................................................................................................................................ 10
3. Results ........................................................................................................................................................... 11
3.1. LC-MRI contrasts and DTI in BASE-II ............................................................................................... 11
3.2. Fractional anisotropy in the LC is higher in older adults, relative to younger adults .......................... 13
4. Discussion ...................................................................................................................................................... 23
4.1. Limitations ........................................................................................................................................... 27
4.2. Conclusions .......................................................................................................................................... 27
5. Supplementary material ................................................................................................................................. 27
5.1. Mean and Radial diffusivity show similar trends to fractional anisotropy. ......................................... 27
Chapter 2: Trajectory of locus coeruleus diffusivity across the lifespan ..................................................................... 41
1. Introduction .................................................................................................................................................... 41
2. Methods.......................................................................................................................................................... 43
2.1. Processing diffusion images ................................................................................................................. 44
2.2. Statistical analyses ................................................................................................................................ 47
3. Results ............................................................................................................................................................ 49
3.1. Diffusivity in the LC between young and older adults ........................................................................ 49
3.2. Diffusivity across the lifespan in the cross-sectional cohort ................................................................ 51
3.3. Differences between each ROI smooth terms for fractional anisotropy and axial diffusivity ............. 59
3.4. Longitudinal trajectories in diffusivity in young and older adults ....................................................... 61
3.5. Comparison of longitudinal trajectories between regions of interest ................................................... 63
4. Discussion ...................................................................................................................................................... 64
4.1. Limitations ........................................................................................................................................... 69
5. Supplementary material ................................................................................................................................. 70
Chapter 3: LC diffusivity may be influenced by sleep deprivation ............................................................................. 86
1. Introduction .................................................................................................................................................... 86
2. Methods.......................................................................................................................................................... 88
2.1. DWI processing .................................................................................................................................... 90
2.2. Statistical analyses ................................................................................................................................ 91
RUNNING HEAD: AGING AND LC DIFFUSIVITY
iv
3. Results ............................................................................................................................................................ 92
3.1. Fractional anisotropy in the LC is higher in older adults, relative to younger adults .......................... 92
4. Discussion ...................................................................................................................................................... 97
4.1. Limitations ......................................................................................................................................... 101
4.2. Conclusions ........................................................................................................................................ 101
5. Supplementary material ............................................................................................................................... 102
Chapter 4: General Discussion ................................................................................................................................... 106
1. Main Finding 1: Fractional anisotropy in the LC of younger adults is lower when compared with older
adults ...................................................................................................................................................................... 107
2. Main Finding 2: Fractional anisotropy in the LC displays atypical non-linear trajectories with aging,
compared with the ascending noradrenergic bundle .............................................................................................. 112
3. Main Finding 3: Sleep deprivation may affect fractional anisotropy in the LC of younger adults more than
older adults ............................................................................................................................................................. 117
4. Concluding Remarks .................................................................................................................................... 121
References .................................................................................................................................................................. 123

RUNNING HEAD: AGING AND LC DIFFUSIVITY
v
List of Tables

Table 1: Demographics .................................................................................................................................................. 4
Table 2: LC-MRI Contrast Sequence Parameters .......................................................................................................... 5
Table 3: Structural MRI Sequence Parameters in Each Study ....................................................................................... 6
Table 4: Diffusion MRI Sequence Parameters in Each Study ....................................................................................... 7
Table 5: Younger adults LC-MRI contrast correlations with confidence intervals ..................................................... 12
Table 6: Older adults LC-MRI contrast correlations with confidence intervals .......................................................... 13
Table 7: BASE-II Fractional Anisotropy Mixed ANOVA .......................................................................................... 14
Table 8: LEMON Fractional Anisotropy Mixed ANOVA .......................................................................................... 15
Table 9: BASE-II Fractional Anisotropy Means, Standard Error, Degrees of Freedom and 95% Confidence Intervals
...................................................................................................................................................................................... 18
Table 10: LEMON Fractional Anisotropy Means, Standard Error, Degrees of Freedom and 95% Confidence
Intervals ........................................................................................................................................................................ 19
Table 11: Younger adults LC-MRI contrast correlations with confidence intervals ................................................... 30
Table 12: Older adults LC-MRI contrast correlations with confidence intervals ........................................................ 31
Table 13: BASE-II Mean Diffusivity Mixed ANOVA ................................................................................................ 32
Table 14: LEMON Mean Diffusivity Mixed ANOVA ................................................................................................ 33
Table 15: BASE-II Radial Diffusivity Mixed ANOVA ............................................................................................... 34
Table 16: LEMON Radial Diffusivity Mixed ANOVA ............................................................................................... 35
Table 17: BASE-II Mean and Radial Diffusivity Means, Standard Error, Degrees of Freedom and 95% Confidence
Intervals ........................................................................................................................................................................ 36
Table 18: LEMON Mean and Radial Diffusivity Means, Standard Error, Degrees of Freedom and 95% Confidence
Intervals ........................................................................................................................................................................ 37
Table 19: BASE-II and LEMON AGExROI Specificity ANOVAs ............................................................................ 38
Table 20: Demographics for the cross-sectional cohort. .............................................................................................. 44
Table 21: Demographics for the longitudinal cohort at each visit. .............................................................................. 44
Table 22: GAM Results for Fractional Anisotropy in the Ascending Noradrenergic Bundle, Locus Coeruleus, and
Frontopontine Tract, Cross-Sectional Cohort .............................................................................................................. 53
Table 23: GAM results for Axial Diffusivity in the Ascending Noradrenergic Bundle, LC, and Frontopontine Tract,
Cross-Sectional Cohort ................................................................................................................................................ 58
Table 24: Marginal Derivatives for Fractional Anisotropy and Age in the Locus Coeruleus, Cross-Sectional Cohort
...................................................................................................................................................................................... 70
Table 25: Marginal Derivatives for Fractional Anisotropy and Age in the Ascending Noradrenergic Bundle, Cross-
Sectional Cohort ........................................................................................................................................................... 70
Table 26: Marginal Derivatives for Fractional Anisotropy and Age in the Frontopontine Tract, Cross-Sectional
Cohort ........................................................................................................................................................................... 70
Table 27: Fractional Anisotropy in the Locus Coeruleus with RAVLT, Cross-Sectional Cohort .............................. 71
Table 28: Fractional Anisotropy in the Ascending Noradrenergic Bundle with RAVLT Delay Scores, Cross-
Sectional Cohort ........................................................................................................................................................... 71
RUNNING HEAD: AGING AND LC DIFFUSIVITY
vi
Table 29: Fractional Anisotropy in the Frontopontine with RAVLT, Cross-Sectional Cohort ................................... 71
Table 30: Marginal Derivatives for Axial Diffusivity and Age in the Locus Coeruleus, Cross-Sectional Cohort ...... 72
Table 31: Marginal Derivatives for Axial Diffusivity and Age in the Ascending Noradrenergic Bundle, Cross-
Sectional Cohort ........................................................................................................................................................... 72
Table 32: Marginal Derivatives for Axial Diffusivity and Age in the Frontopontine Tract, Cross-Sectional Cohort 72
Table 33: Axial Diffusivity in the Locus Coeruleus with RAVLT, Cross-Sectional Cohort ...................................... 73
Table 34: Axial diffusivity in the Ascending Noradrenergic Bundle with RAVLT Delay Scores, Cross-Sectional
Cohort ........................................................................................................................................................................... 73
Table 35: Axial Diffusivity in the Frontopontine with RAVLT, Cross-Sectional Cohort .......................................... 73
Table 36: GAMM Results for Fractional Anisotropy in the LC, Young Adults – Longitudinal Cohort .................... 74
Table 37: GAMM Results for Fractional Anisotropy in LC, Older Adults – Longitudinal Cohort ............................ 74
Table 38: GAMM Results for Fractional Anisotropy in the Ascending Noradrenergic Bundle, Young Adults –
Longitudinal Cohort ..................................................................................................................................................... 75
Table 39: GAMM Results for Fractional Anisotropy in the Ascending Noradrenergic Bundle, Older Adults –
Longitudinal Cohort ..................................................................................................................................................... 75
Table 40: GAMM Results for Fractional Anisotropy in the Frontopontine, Young Adults – Longitudinal Cohort ... 76
Table 41: GAMM Results for Fractional Anisotropy in the Frontopontine, Older Adults – Longitudinal Cohort ..... 76
Table 42: GAMM Results for Axial Diffusivity in the LC, Young Adults – Longitudinal Cohort ............................ 77
Table 43: GAMM Results for Axial Diffusivity in LC, Older Adults – Longitudinal Cohort .................................... 77
Table 44: GAMM Results for Axial Diffusivity in the Ascending Noradrenergic Bundle, Young Adults –
Longitudinal Cohort ..................................................................................................................................................... 78
Table 45: GAMM Results for Axial Diffusivity in the Ascending Noradrenergic Bundle, Older Adults –
Longitudinal Cohort ..................................................................................................................................................... 78
Table 46: GAMM Results for Axial Diffusivity in the Frontopontine, Young Adults – Longitudinal Cohort ........... 79
Table 47: GAMM Results for Axial Diffusivity in the Frontopontine, Older Adults – Longitudinal Cohort ............ 79
Table 48: Linear Mixed Effects Portion of Longitudinal GAMM Results .................................................................. 80
Table 49: GAM Smooth Portion of Longitudinal GAMM Results ............................................................................. 81
Table 50: Linear Mixed Effects Portion of Longitudinal GAMM RAVLT Results .................................................... 83
Table 51: GAM Smooth Portion of Longitudinal GAMM RAVLT Results ............................................................... 84
Table 52: Demographics .............................................................................................................................................. 89
Table 53: Diffusion MRI Sequence Parameters ........................................................................................................... 89
Table 54: SLEEPY Fractional Anisotropy Mixed ANOVA ........................................................................................ 94
Table 55: SLEEPY Fractional Anisotropy Means, Standard Error, Degrees of Freedom and 95% Confidence
Intervals ........................................................................................................................................................................ 95
Table 56: SLEEPY Mean Diffusivity Mixed ANOVA ............................................................................................. 103
Table 57: SLEEPY Radial Diffusivity Mixed ANOVA ............................................................................................ 104
Table 58: SLEEPY Mean Diffusivity Means, Standard Error, Degrees of Freedom and 95% Confidence Intervals105
Table 59: SLEEPY Radial Diffusivity Means, Standard Error, Degrees of Freedom and 95% Confidence Intervals
.................................................................................................................................................................................... 106
RUNNING HEAD: AGING AND LC DIFFUSIVITY
vii
List of Figures

Figure 1: The ROI atlases of the Locus Coeruleus, Noradrenergic Bundle, and Frontopontine Tract .......................... 9
Figure 2: BASE-II Fractional Anisotropy in Left and Right Locus Coeruleus and Noradrenergic Bundles in Younger
and Older Adults .......................................................................................................................................................... 16
Figure 3: LEMON Fractional Anisotropy in Left and Right Locus Coeruleus and Noradrenergic Bundles in Younger
and Older Adults .......................................................................................................................................................... 17
Figure 4: BASE-II Fractional Anisotropy Along the Noradrenergic Bundle .............................................................. 21
Figure 5: LEMON Fractional Anisotropy Along the Noradrenergic Bundle .............................................................. 22
Figure 6: BASE-II NA + Frontopontine ANOVA Contrast Plots ............................................................................... 38
Figure 7: BASE-II LC + Frontopontine ANOVA Contrast Plots ................................................................................ 39
Figure 8: LEMON NA + Frontopontine ANOVA Contrast Plots ............................................................................... 39
Figure 9: LEMON LC + Frontopontine ANOVA Contrast Plots ................................................................................ 40
Figure 10: The LC, Ascending Noradrenergic bundle, and Frontopontine Atlases. ................................................... 46
Figure 11: Fractional Anisotropy Differences between Young and Older Adults in the LC and the Ascending
Noradrenergic Bundle .................................................................................................................................................. 50
Figure 12: GAM Fit and Marginal Derivatives for FA in each Region of Interest ..................................................... 54
Figure 13: GAM Relationship between Fractional Anisotropy and RAVLT Delay Scores in Young and Older
Adults, Cross-Sectional Cohort .................................................................................................................................... 55
Figure 14: GAM Fit and Marginal Derivatives for Axial Diffusivity in each ROI ..................................................... 58
Figure 15: Difference between ROI Smooth Terms across Age in Fractional Anisotropy ......................................... 60
Figure 16: Difference between ROI Smooth Terms across Age in Axial Diffusivity ................................................. 61
Figure 17: Longitudinal Changes in Fractional Anisotropy by Age and Gender ........................................................ 62
Figure 18: Longitudinal Changes in Axial Diffusivity by Age and Gender ................................................................ 63
Figure 19: GAMM Longitudinal Fixed Effects Between Age Groups and 95% Confidence Intervals ...................... 82
Figure 20: GAMM RAVLT Longitudinal Fixed Effects Between Age Groups and 95% Confidence Intervals ........ 85
Figure 21: The ROI atlases of the Locus Coeruleus, Noradrenergic Bundle, and Frontopontine Tract ...................... 92
Figure 22: SLEEPY Brain Fractional Anisotropy in Left and Right Locus Coeruleus and Noradrenergic Bundles in
Younger and Older Adults ........................................................................................................................................... 96
Figure 23: SLEEPY Brain Fractional Anisotropy in Left and Right Locus Coeruleus and Noradrenergic Bundles
between Younger and Older Adults ............................................................................................................................. 97

RUNNING HEAD: AGING AND LC DIFFUSIVITY

Abstract

The locus coeruleus (LC) is essential for arousal and attentional behavior. As the brain’s
primary source for noradrenaline, the health of the LC is paramount to the central nervous
system. Previous studies have identified the LC as the initial site of abnormal tau in the
Alzheimer’s disease progression of pathology. However, in-vivo investigations of the LC have
been difficult given its small size and deep brainstem location. Recently, neuroimaging
techniques like diffusion tensor imaging, have allowed us to study the microstructural integrity
of the LC in-vivo in humans. For example, fractional anisotropy is widely used as a measure of
structural integrity (higher fractional anisotropy typically indicates healthier axons) and has a
strong inverse correlation with axial, radial and mean diffusivities.
The aims of this dissertation were to identify age-related diffusivity of the LC in relation
to the LC ascending noradrenergic bundle, a white matter tract that projects axons from the LC
and terminates in the entorhinal cortex. Specifically, my goal was to first replicate existing
literature reports of increased fractional anisotropy in the LC of older adults, relative to younger
adults (an atypical age-related discovery). Then I investigated if these changes in diffusion are
observed consistently throughout the lifespan. Given the LC ascending noradrenergic bundle
terminates in the entorhinal cortex, I also investigated its relationship to verbal memory. Finally,
I attempted to explore possible underlying mechanisms that may be explain age-related
diffusivity differences in the LC with a sleep-deprived cohort.
In all three studies, I used neuroimaging techniques to identify diffusivity differences in
the LC and the LC ascending noradrenergic bundle in publicly available datasets. Study 1
utilized the Berlin Aging Study-II (N = 301) and the Leipzig Study for Mind-Body-Emotion
RUNNING HEAD: AGING AND LC DIFFUSIVITY
ix
Interactions (N = 220) in attempt to replicate previous LC fractional anisotropy findings by
Langley et al. (2020) in relation to the LC ascending noradrenergic bundle. Results show a
successful replication of LC fractional anisotropy, with atypical age-related findings of younger
adults displaying less fractional anisotropy relative to older adults. However, the LC ascending
noradrenergic bundle displayed expected age-related changes in fractional anisotropy, with
younger adults observed to have more fractional anisotropy than older adults. Additionally, our
control region, the frontopontine tract that runs anterior to the LC, was observed to have no
significant difference in fractional anisotropy between age groups. These findings suggest that
diffusivity in the LC versus the LC ascending noradrenergic bundle may be susceptible to
opposing microstructural changes throughout aging.
In study 2, I expanded the aims of our first study and utilized the Nathan Kline Institute
(NKI) / Rockland dataset (N=369), to identify how diffusivity in the LC and the LC ascending
noradrenergic bundle change throughout the lifespan, from early childhood (6 years old) to late
life (75 years old). Our results replicate aim 1 age-related differences of fractional anisotropy of
the LC and expand support for the hypothesis that diffusivity in the LC may be susceptible to
opposing microstructural changes in relation to the LC ascending noradrenergic bundle.
However, we were not able to replicate differences in fractional anisotropy in the LC ascending
noradrenergic bundle between specific young and older adults age-groups. Yet, striking
differences in axial, radial, and mean diffusivity that demonstrate age-related changes across the
lifespan may be uniquely different in the LC compared to the LC ascending noradrenergic
bundle.
Approximately 169 participants had two additional follow up visits for which I observed
consistent diffusivity trends in each region of interest. Finally, I was not able to replicate ray-
RUNNING HEAD: AGING AND LC DIFFUSIVITY
x
auditory verbal learning test (RAVLT) results identified previously with LC diffusivity but did
identify associations with the LC noradrenergic ascending bundle. Together, these findings
further demonstrate that LC diffusivity is unique in relation to the LC ascending noradrenergic
bundle and frontopontine tracts. Finally, I discuss that using fixed age-ranges to study diffusivity
changes may not be as informative as utilizing lifespan data.
In study 3, I explore what may drive the underlying mechanisms for the observed age-
related differences in diffusivity between the LC and its ascending noradrenergic bundle. For
this, I turned to the Stockholm Sleepy Brain dataset (N = 49), where young and older adults
underwent a single night of sleep deprivation before their MRI scan. The results suggested that
age-related differences in fractional anisotropy may be associated with sleep deprivation.
Specifically, I was only able to replicate the increased fractional anisotropy in the LC of older
adults when comparing sleep deprived participants. We further observed that sleep deprived
young adults, relative to sleep rested young adults, had lower fractional anisotropy in the LC.
But, sleep-deprived older adults did not show any difference in fractional anisotropy, compared
with sleep rested older adults. This suggests that the LC of younger adults may be more
susceptible to sleep deprivation effects than older adults. These findings, although limited in
scope, may provide a framework for a future sleep-deprivation diffusion imaging study that
seeks to replicate our results.
Together, the findings from these three studies provide evidence to suggest that the
microstructural changes in the locus coeruleus are unique and do not adhere to typical age-
related declines in diffusivity as observed with the LC ascending noradrenergic bundle and
frontopontine tract. Additionally, the sleep-related observations support further cause for
replication in larger cohorts, as they have broad implications to how sleep may impact the health
RUNNING HEAD: AGING AND LC DIFFUSIVITY
xi
of the locus coeruleus. Given the role of the LC in arousal and attention, as well as diseases like
Alzheimer’s, the results in these studies provide a framework for further exploration of non-
invasive neuroimaging biomarkers that may provide clinical benefits to the aging population.

RUNNING HEAD: AGING AND LC DIFFUSIVITY

Chapter 1: Age differences in locus coeruleus diffusivity and its
ascending noradrenergic bundle

1. Introduction

The locus coeruleus (LC) is the brain’s primary source for noradrenaline (F. S. Giorgi et
al., 2020; Khanday et al., 2016; Lee et al., 2018), influencing arousal and attention (Aston-Jones
& Waterhouse, 2016; Dahl, Mather, Sander, et al., 2020; Mather, 2020; Mather & Harley, 2016;
McGregor & Siegel, 2010; Sara, 2009). The LC also has widespread cortical projections that are
susceptible to neurodegeneration (Aston-Jones & Waterhouse, 2016; Loizou, 1969; Loughlin et
al., 1982; Morris, McCall, et al., 2020). Notably, the human LC is the primary site of early
abnormal tau pathology (Braak & Del Trecidi, 2015; Liu et al., 2020; Mather & Harley, 2016)
and until recently, in vivo microstructural properties of the LC were mostly unexplored (Edlow et
al., 2016; Edlow et al., 2012; Langley et al., 2020).
Recently, Langley et al. (2020) examined the diffusive properties of the LC using
diffusion-weighted MRI. They observed higher fractional anisotropy in the LC of older adults,
compared with younger adults. Fractional anisotropy is widely used as a measure of structural
integrity (higher fractional anisotropy typically indicates healthier axons) and has a strong
inverse correlation with mean or radial diffusivity (Beck et al., 2021; Bhagat & Beaulieu, 2004;
Kantarci et al., 2017; Kiely et al., 2022; Kochunov et al., 2012). With aging, older adults
typically display lower fractional anisotropy and higher mean diffusivity in white and gray
matter compared with younger adults (Beck et al., 2021; Kantarci, 2014; Lawrence et al., 2021;
D. A. Medina & M. Gaviria, 2008; Rose et al., 2008; Sullivan & Pfefferbaum, 2006; A. N.
Voineskos et al., 2012). In addition, grey matter also typically shows lower fractional anisotropy
RUNNING HEAD: AGING AND LC DIFFUSIVITY
2
and higher mean diffusivity in Alzheimer’s disease (Kantarci, 2014; Rose et al., 2008; Weston et
al., 2015). Thus, Langley’s findings are the opposite of typical white matter age effects.
Given the surprising nature of the increased fractional anisotropy seen in older adults’ LC
compared with younger adults’ LC, we were interested in testing whether these age differences
replicate in larger samples. Using two large datasets (Berlin Aging Study-II, N = 301, (Delius et
al., 2015), and the Leipzig Study for Mind-Body-Emotion Interactions, N = 220, (Babayan et al.,
2019) of healthy young and older adults, we examined whether we could replicate LC fractional
anisotropy findings as reported by Langley, et al.. We also compared fractional anisotropy in the
LC with fractional anisotropy in the ascending noradrenergic bundle, which originates in the LC.
To characterize diffusion properties within the ascending noradrenergic bundle, we relied on a
probabilistic atlas of bilateral ascending noradrenergic fiber bundles originating in the LC and
terminating in the transentorhinal cortex based on data from the Human Connectome Project
(Sun et al., 2020; Tang et al., 2018).
2. Methods

Demographics and MRI sequence information across all datasets can be found in Tables
1 through Table 4. The first dataset we examined is the Berlin Aging Study II (BASE-II)
(Bertram et al., 2014; Delius et al., 2015) from timepoint two. BASE-II information can be found
online (https://www.base2.mpg.de/en). BASE-II participants signed written informed consent
and received monetary compensation for participation. MRI acquisitions were approved by the
ethics committees of the German Psychological Society (Delius et al., 2015). All experiments
were performed in accordance with relevant guidelines and regulations. In short, BASE-II is a
multi-disciplinary and multi-institutional longitudinal study sampling from Berlin’s population.
Because the BASE-II study included LC-MRI contrast measures, we asked whether these
RUNNING HEAD: AGING AND LC DIFFUSIVITY
3
measures were associated with measures of LC and noradrenergic bundle diffusivity. The LC-
MRI index potentially reflects neuromelanin accumulation as a byproduct of NE synthesis.
Hence, it is supposed to indicate functional NE-density within the LC. If a lower LC-MRI
contrast indeed reflects impaired functionality of the LC-NE system, detrimental downstream
effects on pathways connecting the LC to the entorhinal cortex might be expected. Thus, we
expect lower LC-MRI contrast ratios to be associated with lower diffusivity. BASE-II LC-MRI
contrast values were previously quantified in a different study (Dahl et al., 2019). Briefly,
participants completed two LC-sensitive brainstem scans (Table 2). To improve signal-to-noise
ratio, scans were motion corrected and ratios were extracted from both scans, then averaged
across subjects (Dahl et al., 2019). LC location was identified semi-automatically and each
participant’s LC intensity values were extracted and averaged to obtain a reliable estimate (Dahl
et al., 2019).
The second dataset we examined is the Leipzig Study for Mind-Body-Emotion
Interactions (LEMON), for which extensive details can be found elsewhere (Babayan et al.,
2019). The Declaration of Helsinki was followed in this study and the ethics committee at the
medical faculty of the University of Leipzig (reference number 154/13-ff) approved the LEMON
study protocol (Babayan et al., 2019).This cross-sectional study contains both young and older
adults from Leipzig, Germany, and the surrounding area. Subject demographics in each study,
with final N’s per dataset, are displayed in Table 1. We excluded subjects with poor quality
diffusion-weighted scans, poor atlas registration, or missing data in statistical analyses. Poor scan
and registration quality was determined through visual inspection. Tables 2-4 contain LC-MRI
contrast sequence parameters, structural MRI parameters, and diffusion-weighted MRI
parameters across studies, respectively.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
4
Table 1: Demographics
Younger Adults Older Adults p
a

Berlin Aging Study-II (BASE-II)
Age in Years
b
35.90 (3.67) 75 .65 (4.05) <0.001
Sex
c
0.6
Male 39 (67) 154 (63)
Female 19 (33) 89 (37)
Total 58 243
Leipzig Study for Mind-Body-Emotion Interactions (LEMON)
Age in Years 25.10 (3.10) 67.60 (4.70) <0.001
Sex 0.004
Male 105 (70) 35 (49)
Female 44 (30) 36 (51)
Total 149 71
a
Statistical tests performed: chi-square test of independence (for comparisons across sexes and
sleep deprivation conditions); Wilcoxon rank-sum test (for age).
b
Statistics presented: Mean
(SD).
c
Statistics presented: n (% of total)

RUNNING HEAD: AGING AND LC DIFFUSIVITY
5
Table 2: LC-MRI Contrast Sequence Parameters
BASE-II Item Duration or Size
Scanner 3-Tesla Siemens Magnetom
Tim Trio

Head-coil 12-channel
Sequence High-resolution, two-
dimensional T1-weighted
turbo-spin echo (TSE) sequence
aligned perpendicularly to the
plane of the respective
participant’s brainstem
2 * 5.9 minutes
Parameters Repetition Time 600ms
Echo Time 11ms
Inversion Time
Flip Angle 120 °
Bandwidth 287Hz/pixel
FOV 350mm
Slice Thickness 2.5mm isotropic

RUNNING HEAD: AGING AND LC DIFFUSIVITY
6
Table 3: Structural MRI Sequence Parameters in Each Study
Item Duration or Size
Berlin Aging Study-II (BASE-II)
Scanner 3-Tesla Siemens Magnetom Tim
Trio

Head coil 12-channel
Sequence T1-weighted magnetization
prepared gradient-echo (MPRAGE)
9:2 minutes
Parameters Repetition Time 2500ms
Echo Time 4.77ms
Inversion Time 1100ms
Flip Angle 7 °
Bandwidth 140Hz/pixel
FOV 256mm
Slice Thickness 1mm isotropic
Other The LC-contrast scans acquired when participants returned for visit
2. Pre-scan normalize, and 3D distortion correction options were
enabled.
Leipzig Study for Mind-Body-Emotion Interactions (LEMON)
Scanner 3-Tesla Siemens Magnetom Verio
Head coil 32-channel
Sequence Magnetization Prepared 2 Rapid
Acquisition Gradient Echoes
8:22 minutes
Parameters Repetition Time 5000ms
Echo Time 2.92ms
Inversion Time 1/2 700/2500ms
Flip Angle 1/2 4/5°
Bandwidth 240 Hz/pixel
FOV 256mm
Slice Thickness 1mm isotropic
Other For more information, please see Babayan et al., (2019).

RUNNING HEAD: AGING AND LC DIFFUSIVITY
7
Table 4: Diffusion MRI Sequence Parameters in Each Study
Item Duration or Size
Berlin Aging Study-II (BASE-II)
Scanner 3-Tesla Siemens Magnetom Tim Trio
Head coil 12-channel
Sequence Transverse plane and seven volumes
collected without diffusion weighting (b =
0).
12:52 minutes
Parameters Repetition Time 11000ms
Echo Time 98ms
Gradient Directions 60
Diffusion Weighting b = 1,000 s/mm
2

Bandwidth 1628 Hz/pixel
FOV 218mm
Slice Thickness 1.7mm isotropic
Other Generalized auto-calibrating partially parallel acquisitions acceleration
factor = 2 in single-shot, echo-planar imaging. For more details, see
Dahl et al., (2019).
Leipzig Study for Mind-Body-Emotion Interactions (LEMON)
Scanner 3-Tesla Siemens Magnetom Verio
Head coil 32-channel
Sequence Transverse plane and seven volumes
collected without diffusion weighting (b =
0).
9:27 minutes
Parameters Repetition Time 7000ms
Echo Time 80ms
Gradient Directions 60
Diffusion Weighting b = 1,000 s/mm
2

Bandwidth 1502 Hz/pixel
FOV 220mm
Slice Thickness 1.7mm isotropic
Other For more information please see (Babayan et al., 2019).
RUNNING HEAD: AGING AND LC DIFFUSIVITY
8
2.1. DWI Processing

Using University of Southern California’s Laboratory of Neuroimaging (LONI) Pipeline,
we applied FSL’s (v6.3) eddy-current and motion correction, brain extraction tool, and
resampling to isotropic resolution of 2mm
3
(Dinov et al., 2009; Smith et al., 2004). We used
MRtrix (v3.1) to compute fractional anisotropy (FA) and eigenvalue images (Tournier et al.,
2019). With diffusion images, tensors are estimated for each participant and a whole brain
fractional anisotropy image is created. Our atlas of the right and left LC nuclei was obtained
from a LC meta-mask (Dahl et al., 2022) and the right and left noradrenergic bundles from Tang
et al. (2020). As control regions, we utilized the previously defined right and left frontopontine
tracts (Tang et al., 2018), which run along the ventral portion of the pons on either side of the
basilar sulcus, terminating at the pontine nuclei. Figure 1 displays all three ROIs in MNI152
linear, 1mm resolution space.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
9
Figure 1: The ROI atlases of the Locus Coeruleus, Noradrenergic Bundle, and Frontopontine
Tract
Note. Figure 1 displays the locus coeruleus (yellow), noradrenergic bundle (red), and
frontopontine (blue) tracts registered to MNI152 space. The noradrenergic bundle is one
continuous bundle (part of the temporal lobe segment is not pictured).

Both fractional anisotropy and atlas images were registered into MNI152 linear, 1mm
brain space. Using ANTS nonlinear registration (Avants et al., 2008; Sun et al., 2020) the atlases
were then backwarped into individual subject space with nearest neighbor interpolation.
Registration quality was visualized using an in-house MATLAB script (MATLAB ver. R2019a).
After accurate atlas registration to individual subject space was confirmed with visual inspection,
mean and radial diffusion images were created from eigenvalue images in MATLAB with
custom scripts. Atlases were then converted into a binary mask and multiplied by the diffusion
image to provide fractional anisotropy, mean, and radial diffusivity values along the atlases, per
RUNNING HEAD: AGING AND LC DIFFUSIVITY
10
voxel, within the native space. Diffusivity values were then averaged to provide one diffusivity
value per participant within each ROI.
Since the noradrenergic bundle overlaps with a portion of the LC atlas, we conducted an
along-tract analysis of fractional anisotropy of the noradrenergic bundle. 50 equidistant points
were imposed along the noradrenergic bundle as discussed elsewhere (Sun et al., 2020). Each
point was averaged across participants within younger or older adult groups. Though the 50
equidistant points do not represent distinct anatomical regions, based on subject registrations, we
can approximate the first 10 points to represent most of the LC and points 30-50 represent areas
of the entorhinal cortex. Fractional anisotropy along the tract, is also represented visually on the
Y-axis, against each of the 50 equidistant points between younger and older adults, shown in
Figures 4 and 5.
2.2. Statistical Analyses
All statistical analyses were conducted using the R software (Team, 2020) with tidyverse
and various additional packages (Ahlmann-Eltze, 2019; Aust & Barth, 2020; Kassambara; Lenth,
2021; Sjoberg et al., 2021; Wicham, 2017; Wickham, 2016; Xie, 2021). Correlation coefficients
and 95% confidence intervals were used to identify the relationship between LC-MRI contrast
and diffusivity measurements. Diffusivity and fractional anisotropy, mean diffusivity, and radial
diffusivity values in the LC, ascending noradrenergic bundle, and frontopontine tract were
analyzed within each dataset using a 2 × 2 × 3 × 2 factorial design in which age (younger, older)
and gender (female, male) were between-subject factors and ROI (noradrenergic bundle, locus
coeruleus, frontopontine tract) and hemisphere (left, right) were repeated-measures factors.
To clarify the significant interactions of age and the 3-level ROI factors, we conducted
two follow-up ANOVAs with the ROI factor reduced to 2 levels to separately contrast the
RUNNING HEAD: AGING AND LC DIFFUSIVITY
11
control frontopontine tract with each of the other two ROIs (figure and table in supplementary
material). Levene’s tests were used to explore ANOVA assumptions of equal variances. We
report effect sizes using 𝜂̂ !
"
(generalized eta squared) for ANOVA effects and provide 95%
confidence intervals to allow for comparisons across means. Greenhouse-Geisser correction was
automatically computed for ANOVA departures from sphericity. For the along-tract analyses, t-
tests were conducted for fractional anisotropy at each of the 50 equidistant points between
younger and older adults. P values were false-discovery rate adjusted and surviving points of
significant FA differences between age groups are plotted in Figures 4 and 5. Our focus was on
fractional anisotropy, but we include mean and radial diffusivity findings in the supplementary
material. Lastly, to investigate if LC-FA diffusivity is associated with noradrenergic bundle-FA
diffusivity, we conducted Pearson correlations and t-tests for each dataset.
3. Results

3.1. LC-MRI Contrasts and DTI in BASE-II

In the BASE-II dataset, there were no significant differences between young and older
adults’ overall LC-MRI contrast values (Bachman et al., 2021; Dahl et al., 2019). We also did
not observe significant associations between LC-MRI contrast and diffusivity in either the LC or
ascending noradrenergic bundle. Correlation coefficients with FA and 95% confidence intervals
for younger and older adults are displayed in Table 5 and Table 6, respectively. Previous studies
have observed no overall age differences, but spatially confined age differences between caudal
and rostral regions of the LC have been observed with LC-MRI contrast (Bachman et al., 2021;
Dahl et al., 2019).

RUNNING HEAD: AGING AND LC DIFFUSIVITY
12
Table 5: Younger adults LC-MRI contrast correlations with confidence intervals

Variable LC-MRI Contrast

Noradrenergic bundle FA – Left hemisphere -.07
[-.31, .18]

Noradrenergic bundle FA – Right hemisphere -.09
[-.33, .16]

Locus Coeruleus FA – Left hemisphere -.02
[-.27, .23]

Locus Coeruleus FA – Right hemisphere -.06
[-.30, .19]

Frontopontine FA – Left hemisphere .04
[-.21, .29]

Frontopontine FA – Right hemisphere .15
[-.10, .39]

Note. Values in square brackets indicate the 95% confidence interval for each correlation.
The confidence interval is a plausible range of population correlations that could have caused the
sample correlation (Cumming, 2014).

RUNNING HEAD: AGING AND LC DIFFUSIVITY
13
Table 6: Older adults LC-MRI contrast correlations with confidence intervals

Variable LC-MRI Contrast

Noradrenergic bundle FA – Left hemisphere -.03
[-.15, .10]

Noradrenergic bundle FA – Right hemisphere -.08
[-.20, .05]

Locus Coeruleus FA – Left hemisphere .05
[-.08, .17]

Locus Coeruleus FA – Right hemisphere -.10
[-.23, .02]

Frontopontine FA – Left hemisphere -.06
[-.19, .06]

Frontopontine FA – Right hemisphere -.02
[-.14, .11]

Note. Values in square brackets indicate the 95% confidence interval for each correlation.
The confidence interval is a plausible range of population correlations that could have caused the
sample correlation (Cumming, 2014).

3.2. Fractional anisotropy in the LC is higher in older adults, relative to younger adults

Complete ANOVA tables for fractional anisotropy across datasets are displayed in Tables
7-9. Here in the text, we report the significant ANOVA interactions involving Age and ROI. In
the BASE-II and LEMON datasets, we observed a significant interaction of Age x ROI for
fractional anisotropy, F(1.57, 468.27) = 27.18, p < .001, 𝜂̂ !
"
= .033, and F(1.79, 386.34) = 26.07,
p < .001, 𝜂̂ !
"
= 0.035, respectively (Table 7 and Table 8). We also observed a significant 3-way
interaction of Age x ROI x Hemisphere for fractional anisotropy, F(1.62, 483.01) = 6.49, p =
RUNNING HEAD: AGING AND LC DIFFUSIVITY
14
.003, 𝜂̂ !
"
= .05, and F(1.63, 352.15) = 5.50, p = .008, 𝜂̂ !
"
= 0.04, in the BASE-II and LEMON
datasets, respectively.
Table 7: BASE-II Fractional Anisotropy Mixed ANOVA
Variable 𝐹 𝑑𝑓
!
""
𝑑𝑓
#
""
𝑀𝑆𝐸 𝑝 𝜂̂"
#

Age (Younger, Older) 1.05 1 299 0.0
1
.306 .001
Gender (Female, Male) 7.33 1 299 0.0
1
.007 .007
ROI (LC, Noradrenergic Bundle, Frontopontine) 598.17 1.57 468.27 0.0
1
<.001 .426
Hemisphere (Left, Right) 125.52 1 299 0.0
0
<.001 .049
Age × Gender 1.21 1 299 0.0
1
.273 .001
Age × ROI 27.18 1.57 468.27 0.0
1
<.001 .033
Gender × ROI 1.97 1.57 468.27 0.0
1
.151 .002
Age × Hemisphere 0.00 1 299 0.0
0
.980 .000
Gender × Hemisphere 0.17 1 299 0.0
0
.682 .000
ROI × Hemisphere 73.81 1.62 483.01 0.0
0
<.001 .051
Age × Gender × ROI 0.46 1.57 468.27 0.0
1
.584 .001
Age × Gender × Hemisphere 0.46 1 299 0.0
0
.497 .000
Age × ROI × Hemisphere 6.49 1.62 483.01 0.0
0
.003 .005
Gender × ROI × Hemisphere 0.25 1.62 483.01 0.0
0
.729 .000
Age × Gender × ROI × Hemisphere 0.17 1.62 483.01 0.0
0
.802 .000
Note. "GG": applies Greenhouse-Geisser correction to all within-subjects factors.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
15
Table 8: LEMON Fractional Anisotropy Mixed ANOVA
Variable 𝐹 𝑑𝑓
#
!!
𝑑𝑓
"
!!
𝑀𝑆𝐸 𝑝 𝜂̂ !
"

Age (Younger, Older) 0.26 1 216 0.01 .608 .001
Gender (Female, Male) 0.72 1 216 0.01 .396 .002
ROI (LC, Noradrenergic Bundle,
Frontopontine)
761.9
7
1.79 386.34 0.00 < .001 .513
Hemisphere (Left, Right) 412.7
3
1 216 0.00 < .001 .130
Age × Gender 0.38 1 216 0.01 .539 .001
Age × ROI 26.07 1.79 386.34 0.00 < .001 .035
Gender × ROI 2.22 1.79 386.34 0.00 .116 .003
Age × Hemisphere 9.17 1 216 0.00 .003 .003
Gender × Hemisphere 1.69 1 216 0.00 .195 .001
ROI × Hemisphere 308.5
4
1.63 352.15 0.00 < .001 .194
Age × Gender × ROI 0.57 1.79 386.34 0.00 .546 .001
Age × Gender × Hemisphere 4.70 1 216 0.00 .031 .002
Age × ROI × Hemisphere 5.50 1.63 352.15 0.00 .008 .004
Gender × ROI × Hemisphere 1.56 1.63 352.15 0.00 .214 .001
Age × Gender × ROI × Hemisphere 2.58 1.63 352.15 0.00 .088 .002
Note. "GG": applies Greenhouse-Geisser correction to all within-subjects factors.
Tables 9 and 10 report means and 95% confidence intervals for fractional anisotropy for
each ROI between age groups, in each hemisphere. We observed significantly less fractional
anisotropy in the LC and significantly more fractional anisotropy in the noradrenergic bundle of
younger adults compared to older adults, in both the BASE-II and LEMON datasets (Tables 9
and 10; Figures 2 and 3). We observed no significant differences in frontopontine tract fractional
anisotropy between younger and older adults in either BASE-II or LEMON datasets.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
16
Figure 2: BASE-II Fractional Anisotropy in Left and Right Locus Coeruleus and Noradrenergic
Bundles in Younger and Older Adults

Note. Figure 2 displays fractional anisotropy between younger and older adults from the BASE-
II cohort. In the left locus coeruleus (A) and right locus coeruleus (B), we observed lower
fractional anisotropy in younger adults, compared to older adults. In the left noradrenergic
bundle (C) and right noradrenergic bundle (D) we observed higher fractional anisotropy in
younger adults, relative to older adults.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
17
Figure 3: LEMON Fractional Anisotropy in Left and Right Locus Coeruleus and Noradrenergic
Bundles in Younger and Older Adults

Note. Figure 3 displays our BASE-II replicated fractional anisotropy findings between younger
and older adults in the LEMON dataset. In the left locus coeruleus (A) and right locus coeruleus
(B), we observed lower fractional anisotropy in younger adults, compared to older adults. In the
left noradrenergic bundle (C) and right noradrenergic bundle (D) we observed higher fractional
anisotropy in younger adults, relative to older adults.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
18
Table 9: BASE-II Fractional Anisotropy Means, Standard Error, Degrees of Freedom and 95%
Confidence Intervals
BASE-II Locus Coeruleus Noradrenergic
Bundle
Frontopontine Tract
Young
Adult
Older
Adult
Young
Adult
Older
Adult
Young
Adult
Older
Adult
Left Hemisphere
M
a
0.396 0.426 0.421 0.403 0.563 0.567
SE 0.01 0.005 0.004 0.002 0.008 0.004
Lower
CI
0.377 0.417 0.413 0.4 0.547 0.559
Upper
CI
0.416 0.436 0.428 0.407 0.58 0.574
Right Hemisphere
M
a
0.397 0.452 0.356 0.309 0.515 0.522
SE 0.014 0.007 0.004 0.002 0.008 0.004
Lower
CI
0.369 0.439 0.348 0.306 0.5 0.515
Upper
CI
0.424 0.465 0.363 0.313 0.53 0.529
Note. CI = confidence interval,
a
degrees of freedom = 299.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
19
Table 10: LEMON Fractional Anisotropy Means, Standard Error, Degrees of Freedom and 95%
Confidence Intervals
LEMON Locus Coeruleus Noradrenergic
Bundle
Frontopontine Tract
Young
Adult
Older
Adult
Young
Adult
Older
Adult
Young
Adult
Older
Adult
Left Hemisphere
M
a
0.427 0.459 0.445 0.427 0.584 0.582
SE 0.005 0.006 0.003 0.005 0.006 0.008
Lower
CI
0.418 0.446 0.438 0.418 0.572 0.567
Upper
CI
0.437 0.472 0.451 0.436 0.595 0.597
Right Hemisphere
M
a
0.473 0.487 0.365 0.315 0.497 0.503
SE 0.007 0.009 0.003 0.004 0.005 0.007
Lower
CI
0.46 0.47 0.359 0.308 0.487 0.489
Upper
CI
0.486 0.505 0.37 0.323 0.508 0.517
Note. CI = confidence interval,
a
degrees of freedom = 216.

Separate ANOVAs with the “ROI” factor either contrasting the frontopontine tract and
noradrenergic bundle fractional anisotropy or contrasting the frontopontine tract and LC
fractional anisotropy both yielded significant Age X ROI effects (Table 9; Figures 1-4 in
supplementary material), F(299) = 26.16, p < .001, 𝜂̂ !
"
= .024, and F(299) = 9.72, p = .002, 𝜂̂ !
"
=
0.009, respectively in BASE-II. As well as in LEMON datasets, F(216) = 24.34, p < .001, 𝜂̂ !
"
=
.025, and F(216) = 5.23, p = .023, 𝜂̂ !
"
= 0.006, respectively. This indicates that the effects of age
RUNNING HEAD: AGING AND LC DIFFUSIVITY
20
on fractional anisotropy in the LC and noradrenergic bundle each differed significantly from the
control region, in opposite directions.
The BASE-II and LEMON along-tract analyses (Figure 4-5) show effects that are
consistent with the LC and noradrenergic bundle results described above. Along the first 10
points, which approximately represent regions close to the LC, younger adults display
significantly lower fractional anisotropy, relative to older adults. In the remaining tract points,
younger adults had higher fractional anisotropy, relative to older adults, with significant age
differences toward the end of the tract, in the region of the entorhinal cortex.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
21
Figure 4: BASE-II Fractional Anisotropy Along the Noradrenergic Bundle

Note. ns = not significant; ns not assigned shape. Fractional anisotropy differences between
younger and older adults are shown along the noradrenergic bundle. The bundle was divided into
50 equidistant points and mean fractional anisotropy was calculated for each age group at each
point. Younger adults had significantly lower fractional anisotropy in the first 10 points of the
noradrenergic bundles which would correspond to the area of the locus coeruleus. In contrast,
around the entorhinal cortex, younger adults show higher fractional anisotropy, compared to
older adults. * p ≤ .05. ** p ≤ .01. *** p ≤ .001. **** p ≤ .0001. FDR adjusted.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
22
Figure 5: LEMON Fractional Anisotropy Along the Noradrenergic Bundle

Note. ns = not significant; ns not assigned shape. Fractional anisotropy differences between
LEMON younger and older adults are shown along the noradrenergic bundle. The bundle was
divided into 50 equidistant points and mean fractional anisotropy was calculated for each age
group at each point. Younger adults had significantly lower fractional anisotropy in the first 10
points of the noradrenergic bundle, more so in the left than right, which would correspond to the
area of the locus coeruleus. In contrast, younger adults showed higher fractional anisotropy and
significantly greater differences around the entorhinal cortex, compared to older adults. * p ≤ .05.
** p ≤ .01. *** p ≤ .001. **** p ≤ .0001. FDR adjusted.

Finally, we observed significant positive correlations between LC fractional anisotropy
and noradrenergic bundle fractional anisotropy within the BASE-II older adult cohort in the left
and right hemispheres, r(243) = .24, p < .001 and r(243) = .19, p < .001, respectively. However,
only three percent of the variance was accounted for (R
2
adj = 0.03). We did not observe any
significant relationship in the BASE-II young adult cohort. We were also not able to replicate
these findings in the LEMON dataset.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
23
4. Discussion

Unmyelinated neurons and numerous innervations to blood capillaries may expose the
LC to toxins throughout aging (Bekar et al., 2012; Giorgi et al., 2020; Raichle et al., 1975).
During the waking day, the LC has a high constant spiking rate which accumulates oxidative
stress in the mitochondria of LC neurons (Weinshenker, 2018). In addition, excess noradrenaline
not repackaged into synaptic vesicles promotes LC tau pathology (Kang et al., 2020). Existing
evidence also suggests that older adults may be more at risk of these negative side effects of LC
activity due to higher tonic activity levels (Gutchess et al., 2020; Mather, 2021; Weinshenker,
2018).
Fractional anisotropy has been observed to correlate with white matter integrity,
increasing until the age of about 35-40 and decreasing into late life or with disease (Beck et al.,
2021; Kiely et al., 2022; Kochunov et al., 2012; Li et al., 2016). Additionally, mean, and radial
diffusivity are typically negatively correlated with fractional anisotropy (Beaudet et al., 2020;
Beck et al., 2021; Kiely et al., 2022; Li et al., 2016). Here, using two publicly available datasets
(Babayan et al., 2019; Delius et al., 2015), we examined the age-related diffusivity of the LC,
ascending noradrenergic bundle, and, as a control, frontopontine white matter tracts. We
replicated Langley et al.’s (2020) findings of higher fractional anisotropy in the LC in older
adults compared with younger adults, across two large datasets (BASE-II; LEMON).
While fractional anisotropy tended to be higher in older than younger adults within the
LC itself, older adults typically showed lower fractional anisotropy than younger adults along the
noradrenergic bundle white-matter ascending tract, a typical age-related pattern in white matter
(Beck et al., 2021; Medina & Gaviria, 2008; Sibilia et al., 2017; Sullivan & Pfefferbaum, 2006;
Voineskos et al., 2012). The lack of associations observed in our datasets between LC fractional
RUNNING HEAD: AGING AND LC DIFFUSIVITY
24
anisotropy and noradrenergic bundle fractional anisotropy may suggest these two regions are
affected by aging independently.
In the BASE-II and LEMON datasets, age differences in the noradrenergic bundle
contrasted with a lack of age differences in the right and left control white-matter frontopontine
tracts, suggesting that the age effects in the noradrenergic ascending tract reflect more than just a
global change in white matter. Thus, together, these data indicate that diffusivity properties of
the LC and its ascending noradrenergic tract are affected by aging in opposite ways. Our findings
of age differences in fractional anisotropy in the LC and its ascending tracts extend a growing set
of observations of age differences in the structure of the LC in aging (Brickman et al., 2012;
Chen et al., 2014; Chu et al., 2021; Dahl et al., 2021; Dahl, Mather, Werkle-Bergner, et al., 2020;
Langley et al., 2020; Morris, Tan, et al., 2020; Sun et al., 2019).
To date, most in vivo findings of LC structure have relied on LC-MRI sequences that
show a cross-sectional increase in LC-neuromelanin sensitive contrast from young adulthood
until around age 57, at which point it levels off or declines (Liu et al., 2019), potentially
suggesting a gradual accumulation of neuromelanin followed by noradrenergic degeneration.
One of the two data sets we examined (BASE-II) included neuromelanin-sensitive scans. While
LC-MRI contrast has been validated to correlate with the location of neuromelanin (Keren et al.,
2015), it is not yet entirely clear what factors contribute to currently employed LC-MRI contrast
measures. Current hypotheses include presence of macromolecules (Priovoulos et al., 2020),
density of water protons (Watanabe et al., 2019), and the relative proportion of macromolecular
protons to free water protons (Trujillo et al., 2019).
In the BASE-II dataset, there were no significant correlations between LC-MRI contrast
from those scans and diffusivity measures from the LC or noradrenergic bundle. This suggests
RUNNING HEAD: AGING AND LC DIFFUSIVITY
25
that the diffusivity differences reflect different structural changes than the LC-sensitive scans.
An important future research objective should be to examine the relationship between LC
diffusivity measures and cognition, or markers of brain health, as has been done for LC-MRI
contrast (Clewett et al., 2016; Dahl et al., 2019; Langley et al., 2020). One initial study along
these lines found that medial and radial diffusivity in the LC-thalamus tract was correlated with
memory performance in an older cohort (Langley et al., 2021).
Our results raise the question of what properties of the LC are changing to lead its tissue
to show higher fractional anisotropy with age. One possibility could be an increase in
inflammation that restricts fluid flow, as animal research has demonstrated that increases in
microglial density affect diffusivity, as measured using an orientation dispersion index (Yi et al.,
2019). Another possibility is that the age differences in LC diffusivity relate to age differences
in LC tonic activity levels. Although still an open question, various findings suggest that the LC
is more tonically active in aging (Mather, 2020; Weinshenker, 2018). Age differences in tonic
levels of LC could contribute to differences in diffusivity as neuronal activity increases neuronal
volume, while shrinking the volume of the surrounding fluid-filled spaces (Abe et al., 2017;
Iwasa et al., 1980; Le Bihan et al., 2006; Nunes et al., 2021; Svoboda & Syková, 1991; Tirosh &
Nevo, 2013).
Mean and radial diffusivity in the LC also showed some age differences (results and
tables provided in the supplementary material), although not as pronounced as fractional
anisotropy. In the BASE-II dataset, mean diffusivity in the LC was significantly higher in
younger adults, compared to older adults. In the LEMON dataset, mean diffusivity was
significantly higher in the left LC of younger adults, compared to older adults. Though the cause
for these laterality effects is not known, the BASE-II dataset is composed of mostly older adults,
RUNNING HEAD: AGING AND LC DIFFUSIVITY
26
while the LEMON has more younger adults. Given the LC’s proximity to the fourth ventricle,
older adults may be susceptible to neurodegeneration within the LC as well as partial volume
effects (Langley et al., 2020; Liu et al., 2017; Sun et al., 2020).
Because the noradrenergic bundle overlaps with the LC atlas, we conducted an along-
tract analysis for the noradrenergic bundle fractional anisotropy. As expected, we observed
significantly lower FA in the first 10 points of the noradrenergic bundle, which anatomically
approximately represent regions of the locus coeruleus, in younger adults compared with older
adults. Changes in radial diffusivity along the noradrenergic bundle of cognitively impaired older
adults from the Alzheimer's Disease Neuroimaging Initiative were significantly greater,
compared to healthy controls, around the area of the LC and again as the tract approached the
hippocampus (Sun et al., 2020).
While most studies comparing diffusivity in younger and older adults focus on white
matter, a growing number of studies have started to examine diffusivity differences in grey
matter in cortical and subcortical nuclei. Patients with Alzheimer’s disease generally show less
fractional anisotropy and greater mean diffusivity than age-matched healthy adults (Weston et
al., 2015). However, studies following people with autosomal dominant familial Alzheimer’s
disease have found increased mean diffusivity in grey matter regions during the pre-symptomatic
period, and older adults with significant memory decline show lower diffusivity in the posterior
cingulate/precuneus region (Jacobs et al., 2013). As Langley et al., suggested, age-related LC
degeneration may result in restricted diffusion within older adults (Langley et al., 2020).
Fractional anisotropy also shows a positive correlation with age in the caudate, putamen and
globus pallidus in a healthy cohort aged 10-52 (Pal et al., 2011). Thus, the LC may not be the
only brain region showing lower fractional anisotropy in older adults.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
27
4.1. Limitations

Crossing fibers may indicate opposite or unexpected relationships with diffusivity values
that may be related to our unexpected findings (Lee et al., 2015; Oouchi et al., 2007). Despite the
limitations of DTI, it remains a valuable tool that may help us to better understand the LC in-vivo
within humans. In general, our datasets were comprised of younger and older adults that had no
neurological or known sleep disorders and may not reflect the general aging population. We also
did not examine axial diffusivity. Due to partial volume constraints, the locus coeruleus ROI may
be contaminated by white matter and CSF (given the position near the 4
th
ventricle). However,
given the opposite findings in the ascending white matter tract, we were still able to extract
meaningful signal.
4.2. Conclusions

In this study, we identified unique associations of LC diffusivity in the context of healthy
adults across two different data sets. We consistently observed lower fractional anisotropy in the
locus coeruleus of younger adults, compared to older adults but higher fractional anisotropy in
the ascending noradrenergic bundle of younger adults, compared to older adults. Fractional
anisotropy is a measurement of structural integrity, and these age findings add to a growing
literature highlighting age-related differences involving the locus coeruleus. To our knowledge,
this is the first study to compare diffusivity differences in-vivo in the locus coeruleus versus
noradrenergic bundle.
5. Supplementary Material

5.1. Mean and Radial diffusivity show similar trends to fractional anisotropy.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
28
LC-MRI with mean or radial diffusivity correlation coefficients and 95% confidence
intervals for younger and older adults are displayed in Tables 11 and 12, respectively. Like FA,
we found no significant associations between LC-MRI and mean or radial diffusivity. Complete
ANOVA tables for mean and radial diffusivity across datasets are displayed in Tables 3-6. We
observed significant Age x ROI interactions in the BASE-II dataset for mean and radial
diffusivities, F(1.28, 380.26) = 38.34, p < .001, 𝜂̂ !
"
= .05, and F(1.30, 387.03) = 44.29, p < .001,
𝜂̂ !
"
= .057, respectively (Table 13 and Table 15). We also observed significant Age x ROI x
Hemisphere interactions for mean and radial diffusivities in the BASE-II dataset, F(1.40, 414.51)
= 24.32, p < .001, 𝜂̂ !
"
= .15, and F(1.42, 420.66) = 23.81, p < .001, 𝜂̂ !
"
= .016, respectively.
Similarly, we observed significant Age x ROI interactions in the LEMON dataset for
mean and radial diffusivities, F(1.24, 267.66) = 34.65, p < .001, 𝜂̂ !
"
= .064, and F(1.31, 282.94) =
41.16, p < .001, 𝜂̂ !
"
= .073, respectively (Table 14 and Table 16). We also observed significant
Age x ROI x Hemisphere interactions for mean and radial diffusivities in the LEMON dataset,
F(1.41, 304.60) = 28.92, p < .001, 𝜂̂ !
"
= .026, and F(1.46, 314.29) = 28.31, p < .001, 𝜂̂ !
"
= .026,
respectively. Unlike the BASE-II dataset, we observed significant Age x Gender x ROI x
Hemisphere interactions within LEMON mean diffusivity F(1.41, 304.60) = 4.74, p = .019, 𝜂̂ !
"
=
.004, and radial diffusivity, F(1.46, 314.29) = 5.22, p = .012, 𝜂̂ !
"
= .005.
Table 17 and Table 18 report BASE-II and LEMON means and 95% confidence intervals
for mean and radial diffusivity within each ROI between young and older adults in each
hemisphere. We observed varying results between BASE-II and LEMON diffusivities within
ROIs. In BASE-II, mean diffusivity in the left locus coeruleus was not significantly different
between young and older adults, but radial diffusivity was. BASE-II mean and radial diffusivities
in the right locus coeruleus were significantly higher in younger adults, compared to older adults.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
29
In the left and right noradrenergic bundles, both mean and radial diffusivities were significantly
lower in younger adults, compared to older adults. We observed no significant difference in
mean and radial diffusivities between age groups in the BASE-II frontopontine tract.
In the LEMON dataset, we observed mean and radial diffusivity in the left locus
coeruleus to be significantly higher in younger adults, compared to older adults. We did not
observe any statistically significant differences between young or older adult mean and radial
diffusivities in the right locus coeruleus. Younger adults, compared to older adults, had
significantly lower mean and radial diffusivities in the left and right noradrenergic bundles.
Except for significantly higher mean diffusivity in the right frontopontine tract of younger adults,
compared to older adults, no other significant mean or radial differences were observed in the
frontopontine tract between age groups.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
30
Table 11: Younger adults LC-MRI contrast correlations with confidence intervals
Variable LC-MRI Contrast

Noradrenergic bundle MD – Left hemisphere .04
[-.21, .29]

Noradrenergic bundle MD – Right hemisphere .19
[-.06, .42]

Noradrenergic bundle RD – Left hemisphere .10
[-.16, .34]

Noradrenergic bundle RD – Right hemisphere .19
[-.06, .42]

Locus Coeruleus MD – Left hemisphere -.01
[-.26, .24]

Locus Coeruleus MD – Right hemisphere .05
[-.21, .29]

Locus Coeruleus RD – Left hemisphere -.01
[-.26, .24]

Locus Coeruleus RD – Right hemisphere .05
[-.20, .30]

Frontopontine MD – Left hemisphere .11
[-.15, .35]

Frontopontine MD – Right hemisphere -.02
[-.27, .23]

Frontopontine RD – Left hemisphere .02
[-.23, .27]

Frontopontine RD – Right hemisphere -.10
[-.34, .16]

Note. Values in square brackets indicate the 95% confidence interval for each correlation.
The confidence interval is a plausible range of population correlations that could have caused the
sample correlation (Cumming, 2014).

RUNNING HEAD: AGING AND LC DIFFUSIVITY
31
Table 12: Older adults LC-MRI contrast correlations with confidence intervals
Variable LC-MRI Contrast

Noradrenergic bundle MD – Left hemisphere .06
[-.06, .19]

Noradrenergic bundle MD – Right hemisphere .12
[-.00, .25]

Noradrenergic bundle RD – Left hemisphere .06
[-.06, .19]

Noradrenergic bundle RD – Right hemisphere .12
[-.00, .24]

Locus Coeruleus MD – Left hemisphere -.02
[-.15, .11]

Locus Coeruleus MD – Right hemisphere .10
[-.02, .23]

Locus Coeruleus RD – Left hemisphere -.02
[-.15, .10]

Locus Coeruleus RD – Right hemisphere .11
[-.02, .23]

Frontopontine MD – Left hemisphere .02
[-.11, .14]

Frontopontine MD – Right hemisphere .04
[-.08, .17]

Frontopontine RD – Left hemisphere .06
[-.07, .18]

Frontopontine RD – Right hemisphere .01
[-.11, .14]

Note. Values in square brackets indicate the 95% confidence interval for each correlation.
The confidence interval is a plausible range of population correlations that could have caused the
sample correlation (Cumming, 2014).

RUNNING HEAD: AGING AND LC DIFFUSIVITY
32
Table 13: BASE-II Mean Diffusivity Mixed ANOVA
Variable 𝐹 𝑑𝑓
#
!!
𝑑𝑓
"
!!
𝑀𝑆𝐸 𝑝 𝜂̂ !
"

Age (Younger, Older) 7.11 1 297 0.00 .008 .00
7
Gender (Female, Male) 7.37 1 297 0.00 .007 .00
7
ROI (LC, Noradrenergic Bundle,
Frontopontine)
1,376.7
5
1.28 380.2
6
0.00 <.00
1
.65
5
Hemisphere (Left, Right) 78.12 1 297 0.00 <.00
1
.02
5
Age × Gender 1.44 1 297 0.00 .231 .00
1
Age × ROI 38.34 1.28 380.2
6
0.00 <.00
1
.05
0
Gender × ROI 9.12 1.28 380.2
6
0.00 .001 .01
2
Age × Hemisphere 14.04 1 297 0.00 <.00
1
.00
5
Gender × Hemisphere 0.18 1 297 0.00 .671 .00
0
ROI × Hemisphere 217.13 1.40 414.5
1
0.00 <.00
1
.12
2
Age × Gender × ROI 0.02 1.28 380.2
6
0.00 .939 .00
0
Age × Gender × Hemisphere 1.02 1 297 0.00 .313 .00
0
Age × ROI × Hemisphere 24.32 1.40 414.5
1
0.00 <.00
1
.01
5
Gender × ROI × Hemisphere 0.22 1.40 414.5
1
0.00 .722 .00
0
Age × Gender × ROI ×
Hemisphere
0.14 1.40 414.5
1
0.00 .789 .00
0

RUNNING HEAD: AGING AND LC DIFFUSIVITY
33
Table 14: LEMON Mean Diffusivity Mixed ANOVA
Variable 𝐹 𝑑𝑓
#
!!
𝑑𝑓
"
!!
𝑀𝑆𝐸 𝑝 𝜂̂ !
"

Age (Younger, Older) 6.18 1 216 0.00 .014 .008
Gender (Female, Male) 0.01 1 216 0.00 .907 .000
ROI (LC, Noradrenergic Bundle,
Frontopontine)
2,354.77 1.24 267.66 0.00 <.001 .823
Hemisphere (Left, Right) 61.38 1 216 0.00 <.001 .026
Age × Gender 0.00 1 216 0.00 .967 .000
Age × ROI 34.65 1.24 267.66 0.00 <.001 .064
Gender × ROI 4.69 1.24 267.66 0.00 .024 .009
Age × Hemisphere 55.04 1 216 0.00 <.001 .023
Gender × Hemisphere 3.31 1 216 0.00 .070 .001
ROI × Hemisphere 314.66 1.41 304.60 0.00 <.001 .224
Age × Gender × ROI 1.10 1.24 267.66 0.00 .309 .002
Age × Gender × Hemisphere 3.66 1 216 0.00 .057 .002
Age × ROI × Hemisphere 28.92 1.41 304.60 0.00 <.001 .026
Gender × ROI × Hemisphere 1.33 1.41 304.60 0.00 .260 .001
Age × Gender × ROI × Hemisphere 4.74 1.41 304.60 0.00 .019 .004

RUNNING HEAD: AGING AND LC DIFFUSIVITY
34
Table 15: BASE-II Radial Diffusivity Mixed ANOVA
Variable 𝐹 𝑑𝑓
#
!!
𝑑𝑓
"
!!
𝑀𝑆𝐸 𝑝 𝜂̂ !
"

Age (Younger, Older) 3.52 1 297 0.00 .062 .00
3
Gender (Female, Male) 6.16 1 297 0.00 .014 .00
6
ROI (LC, Noradrenergic Bundle,
Frontopontine)
992.1
7
1.30 387.0
3
0.00 <.00
1
.57
5
Hemisphere (Left, Right) 145.0
9
1 297 0.00 <.00
1
.04
7
Age × Gender 2.09 1 297 0.00 .150 .00
2
Age × ROI 44.29 1.30 387.0
3
0.00 <.00
1
.05
7
Gender × ROI 7.10 1.30 387.0
3
0.00 .004 .01
0
Age × Hemisphere 13.56 1 297 0.00 <.00
1
.00
5
Gender × Hemisphere 0.16 1 297 0.00 .690 .00
0
ROI × Hemisphere 184.3
2
1.42 420.6
6
0.00 <.00
1
.11
1
Age × Gender × ROI 0.09 1.30 387.0
3
0.00 .829 .00
0
Age × Gender × Hemisphere 0.79 1 297 0.00 .376 .00
0
Age × ROI × Hemisphere 23.81 1.42 420.6
6
0.00 <.00
1
.01
6
Gender × ROI × Hemisphere 0.28 1.42 420.6
6
0.00 .677 .00
0
Age × Gender × ROI ×
Hemisphere
0.26 1.42 420.6
6
0.00 .695 .00
0

RUNNING HEAD: AGING AND LC DIFFUSIVITY
35
Table 16: LEMON Radial Diffusivity Mixed ANOVA
Variable 𝐹 𝑑𝑓
#
!!
𝑑𝑓
"
!!
𝑀𝑆𝐸 𝑝 𝜂̂ !
"

Age (Younger, Older) 4.62 1 216 0.00 .033 .00
6
Gender (Female, Male) 0.24 1 216 0.00 .624 .00
0
ROI (LC, Noradrenergic Bundle,
Frontopontine)
1,499.7
2
1.31 282.9
4
0.00 <.00
1
.74
1
Hemisphere (Left, Right) 134.75 1 216 0.00 <.00
1
.06
0
Age × Gender 0.42 1 216 0.00 .517 .00
1
Age × ROI 41.16 1.31 282.9
4
0.00 <.00
1
.07
3
Gender × ROI 3.49 1.31 282.9
4
0.00 .051 .00
7
Age × Hemisphere 50.04 1 216 0.00 <.00
1
.02
3
Gender × Hemisphere 3.17 1 216 0.00 .076 .00
2
ROI × Hemisphere 329.55 1.46 314.2
9
0.00 <.00
1
.23
4
Age × Gender × ROI 1.00 1.31 282.9
4
0.00 .338 .00
2
Age × Gender × Hemisphere 4.58 1 216 0.00 .034 .00
2
Age × ROI × Hemisphere 28.31 1.46 314.2
9
0.00 <.00
1
.02
6
Gender × ROI × Hemisphere 1.41 1.46 314.2
9
0.00 .244 .00
1
Age × Gender × ROI ×
Hemisphere
5.22 1.46 314.2
9
0.00 .012 .00
5

RUNNING HEAD: AGING AND LC DIFFUSIVITY
36
Table 17: BASE-II Mean and Radial Diffusivity Means, Standard Error, Degrees of Freedom
and 95% Confidence Intervals
BASE-II Locus Coeruleus Noradrenergic Bundle Frontopontine Tract
Young
Adult
Older
Adult
Young
Adult
Older
Adult
Young
Adult
Older
Adult
Mean Diffusivity, Left Hemisphere
M
a
0.00148 0.00143 0.00079 0.00086 0.00071 0.00073
SE 3.00E-05 2.00E-05 1.00E-05 0 1.00E-05 0
Lower CI 0.00142 0.0014 0.00078 0.00085 7.00E-04 0.00072
Upper CI 0.00154 0.00146 8.00E-04 0.00086 0.00073 0.00073
Mean Diffusivity, Right Hemisphere
M
a
0.00144 0.00136 0.00098 0.00126 0.00069 0.00069
SE 4.00E-05 2.00E-05 2.00E-05 1.00E-05 1.00E-05 0
Lower CI 0.00136 0.00132 0.00094 0.00124 0.00067 0.00068
Upper CI 0.00151 0.00139 0.00102 0.00128 0.00071 7.00E-04
Radial Diffusivity, Left Hemisphere
M
a
0.00117 0.00109 6.00E-04 0.00066 0.00045 0.00046
SE 3.00E-05 2.00E-05 1.00E-05 0 1.00E-05 0
Lower CI 0.00111 0.00106 0.00059 0.00066 0.00044 0.00045
Upper CI 0.00124 0.00113 0.00061 0.00067 0.00047 0.00047
Radial Diffusivity , Right Hemisphere
M
a
0.00114 0.00103 0.00081 0.0011 0.00048 0.00048
SE 4.00E-05 2.00E-05 2.00E-05 1.00E-05 1.00E-05 0
Lower CI 0.00107 0.00099 0.00077 0.00107 0.00046 0.00047
Upper CI 0.00122 0.00107 0.00085 0.00112 5.00E-04 0.00049
Note. CI = confidence interval.
a
degrees of freedom = 297.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
37
Table 18: LEMON Mean and Radial Diffusivity Means, Standard Error, Degrees of Freedom
and 95% Confidence Intervals
LEMON Locus Coeruleus Noradrenergic Bundle Frontopontine Tract
Young
Adult
Older
Adult
Young
Adult
Older
Adult
Young
Adult
Older
Adult
Mean Diffusivity, Left Hemisphere
M
a
0.0015 0.00142 0.00076 0.00079 0.00073 0.00072
SE 2.00E-05 2.00E-05 0 0 1.00E-05 1.00E-05
Lower CI 0.00146 0.00138 0.00076 0.00078 0.00072 0.00071
Upper CI 0.00153 0.00147 0.00077 8.00E-04 0.00074 0.00073
Mean Diffusivity, Right Hemisphere
M
a
0.00135 0.00135 9.00E-04 0.00113 0.00073 0.00071
SE 2.00E-05 2.00E-05 1.00E-05 1.00E-05 0 1.00E-05
Lower CI 0.00132 0.00131 0.00088 0.0011 0.00073 7.00E-04
Upper CI 0.00139 0.0014 0.00092 0.00115 0.00074 0.00072
Radial Diffusivity, Left Hemisphere
M
a
0.00115 0.00106 0.00056 0.00059 0.00045 0.00045
SE 2.00E-05 2.00E-05 0 0 1.00E-05 1.00E-05
Lower CI 0.00112 0.00101 0.00056 0.00058 0.00043 0.00043
Upper CI 0.00118 0.0011 0.00057 6.00E-04 0.00046 0.00046
Radial Diffusivity, Right Hemisphere
M
a
0.00101 0.00099 0.00073 0.00096 5.00E-04 0.00049
SE 2.00E-05 2.00E-05 1.00E-05 1.00E-05 1.00E-05 1.00E-05
Lower CI 0.00097 0.00094 0.00071 0.00094 0.00049 0.00047
Upper CI 0.00104 0.00103 0.00075 0.00099 0.00051 5.00E-04
Note. CI = confidence interval.
a
degrees of freedom = 216.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
38
Table 19: BASE-II and LEMON AGExROI Specificity ANOVAs
BASE-II 𝐹 df 𝑀𝑆𝐸 𝑝 𝜂̂ !
"

Age × ROI (LC + FPT) 9.72 299 0.01 .002 .009
Age × ROI (NA + FPT) 26.16 299 0.00 < .001 .024
LEMON 𝐹 df 𝑀𝑆𝐸 𝑝 𝜂̂ !
"

Age × ROI (LC + FPT) 5.23 216 0.00 .023 .006
Age × ROI (NA + FPT) 24.34 216 0.00 < .001 .025

Figure 6: BASE-II NA + Frontopontine ANOVA Contrast Plots

RUNNING HEAD: AGING AND LC DIFFUSIVITY
39
Figure 7: BASE-II LC + Frontopontine ANOVA Contrast Plots

Figure 8: LEMON NA + Frontopontine ANOVA Contrast Plots

RUNNING HEAD: AGING AND LC DIFFUSIVITY
40

Figure 9: LEMON LC + Frontopontine ANOVA Contrast Plots

RUNNING HEAD: AGING AND LC DIFFUSIVITY
41
Chapter 2: Trajectory of locus coeruleus diffusivity across the
lifespan

1. Introduction

Non-invasive diffusion MRI can shed light on the structural integrity of axons (Beck et
al., 2021; Bhagat & Beaulieu, 2004; Kochunov et al., 2012; SH et al., 2012; Weston et al., 2015;
Zuurbier et al., 2013). Recent investigations into the diffusivity of the locus coeruleus (LC)
yielded atypical age-related observations in fractional anisotropy (Langley et al., 2020; Porat et
al., 2022). Specifically, older adults were observed with higher fractional anisotropy in the LC,
relative to younger adults (Langley et al., 2020). A high degree of fractional anisotropy is
typically inferred as better structural integrity because water molecules are restricted to diffuse
along anatomical paths, like that of an axon (Soares et al., 2013).
Studies have also found positive associations between fractional anisotropy and areas of
dense myelination, suggesting fractional anisotropy is useful for inferring structural integrity
with aging (Chang et al., 2017; Friedrich et al., 2020). In contrast, higher axial diffusivity, which
flows along the main axis of diffusion, is negatively correlated with fractional anisotropy as it
increases with aging (Acosta-Cabronero et al., 2010; Soares et al., 2013). Fractional anisotropy
typically peaks around midlife before declining into old age (Beck et al., 2021; Ethell, 2014; H.
I. Jacobs et al., 2013; Kiely et al., 2022).
In a previous study (see Chapter 1), we replicated these findings in two large data sets
and compared the diffusivity of the LC with its ascending noradrenergic bundle that projects
from the LC and terminates in the transentorhinal cortex (Porat et al., 2022). Unlike the LC, the
ascending noradrenergic bundle followed typical age related declines in fractional anisotropy.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
42
These observations suggest that aging affects the LC’s diffusivity properties differently than the
ascending noradrenergic bundle’s diffusivity properties.
Using diffusion imaging to understand the microstructural properties of the LC is
important for several reasons. First, the LC is the primary site of noradrenaline in the brain,
helping maintain arousal and attention (Clewett et al., 2018; Edlow et al., 2012; Mather, 2020,
2021; McGregor & Siegel, 2010). Second, diffusion imaging may provide a quicker alternative
to measuring structural integrity of the LC relative to LC-specific MRI contrast-ratio scans and
post-mortem histologies of cell-density (Bachman et al., 2021; Keren et al., 2009, 2015). Studies
indicate that higher cell-density and LC contrast is associated with better health and memory
performance (Bachman et al., 2021; Dahl et al., 2019, 2022; Mather, 2020). Lastly, the LC is
observed to be one of the earliest sites of abnormal tau pathology in Alzhimer’s progression
(Braak et al., 2011; Chalermpalanupap et al., 2017; Mather & Harley, 2016). Expanding our
understanding of the LC with diffusion MRI may provide clinical insights when investigating
aging and Alzheimer’s disease.
Here, we aim to further explore age differences in the diffusivity in the LC with the
Enhanced Nathan Kline Institute Rockland Sample data set (NKI) (Nooner et al., 2012). The
NKI data set includes both cross-sectional and longitudinal cohorts ranging from early childhood
to late life. The NKI data provides an opportunity to explore how diffusivity varies across the
lifespan and yearly follow up visits. We investigated both cross-sectional (N = 328) and
longitudinal data (N = 169) for participants with 3 separate visits that included diffusion imaging
and Rey-Auditory Verbal Learning Testing (RAVLT) delay scores. RAVLT delay scores have
been associated with LC integrity and Alzheimer’s dementia (Dahl et al., 2019, 2022; Moradi et
al., 2016).
RUNNING HEAD: AGING AND LC DIFFUSIVITY
43
Using generalized additive models, we explored the nonlinear relationship between
diffusivity and age (Pedersen et al., 2019; Sørensen, Brandmaier, et al., 2021; Wood, 2011).
Because a subset of our cohort had follow up MRI visits, we expanded our analyses with
generalized additive mixed models to explore longitudinal changes in diffusivity and cognition as
well. Our findings support the notion that diffusivity in the LC does not adhere to typical age-
related degeneration and may have different microstructural properties than that of the ascending
noradrenergic bundle and frontopontine tracts.
2. Methods

We examined the publicly available NKI data set (Nooner et al., 2012), available upon
request: (http://fcon_1000.projects.nitrc.org/indi/enhanced/access.html).Diffusion MRI sequences are
publicly available here: (http://fcon_1000.projects.nitrc.org/indi/pro/eNKI_RS_TRT/DIff_137.pdf).
Written informed consent was obtained for all participants and/or their legal guardians, including
consent for public data distribution (Nooner et al., 2012). Briefly, the NKI data set aims to
advance developmental changes across the lifespan. The data set contains both longitudinal and
cross-sectional data for over 1000 participants between the ages of 6 and 85 (Nooner et al.,
2012).
Participants underwent several different MRI scans and were administered behavioral,
cognitive, and physiological tests. In our analyses, there is some overlap with participants
between the cross-sectional and longitudinal data sets. Demographics for participants in the cross
sectional and longitudinal cohorts are found in Table 20 and Table 21, respectively.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
44
Table 20: Demographics for the cross-sectional cohort.

N
Early
Childhood,
N = 29
1

Adolescent,
N = 77
1

Young
Adult, N =
67
1

Middle
Age, N =
66
1

Older
Adult, N =
89
1

p-value
2

Age 328 8 (8, 9) 14 (12, 15) 23 (21, 25) 44 (34, 50) 69 (63, 78) <0.001
Sex 328

<0.001
Female

17 (59%) 31 (40%) 30 (45%) 49 (74%) 51 (57%)

Male

12 (41%) 46 (60%) 37 (55%) 17 (26%) 38 (43%)

1
Median (IQR); n (%).
2
Kruskal-Wallis rank sum test; Pearson's Chi-squared test

Table 21: Demographics for the longitudinal cohort at each visit.

N
Young Adult,
N = 74
1

Older Adult,
N = 95
1

p-value
2

Baseline
Age 169 12 (9, 14) 56 (48, 64) <0.001
Sex 169

<0.001
Female

28 (38%) 64 (67%)

Male

46 (62%) 31 (33%)

Timepoint 2
Age 169 13 (11, 15) 57 (50, 66) <0.001
Sex 169

<0.001
Female

28 (38%) 64 (67%)

Male

46 (62%) 31 (33%)

Timepoint 3
Age 169 14 (12, 16) 58 (50, 66) <0.001
Sex 169

<0.001
Female

28 (38%) 64 (67%)

Male

46 (62%) 31 (33%)

1
Median (IQR); n (%).
2
Wilcoxon rank sum test; Pearson's Chi-squared test

2.1. Processing Diffusion Images

Both NKI cross-sectional and longitudinal data sets follow the same diffusion imaging
protocols available here:
(http://fcon_1000.projects.nitrc.org/indi/pro/eNKI_RS_TRT/DIff_137.pdf). Preprocessing
diffusion images was carried out on each subject, at every visit, using the University of Southern
California’s Laboratory of Neuroimaging (LONI) Pipeline (Dinov et al., 2009). FSL’s (v6.3)
RUNNING HEAD: AGING AND LC DIFFUSIVITY
45
eddy current and motion and distortion correction, brain extraction, and resampling to isotropic
resolution of 2mm
3
(Smith et al., 2004) were applied. To compute diffusion tensors, we
implemented MRtrix (v3.1) dwi2tensor (Tournier et al., 2019). Dahl’s LC meta-mask was
selected to obtain diffusion values from the left and right LC, for its high anatomical accuracy
(Dahl et al., 2019, 2022). The diffusion values of the ascending noradrenergic bundle were
obtained using a publicly available atlas segmented from the Human Connectom data set (Sun et
al., 2020). The frontopontine brainstem tract was used as our control region given its ventral
location to the LC within the pons, and we used a publicly available atlas to delineate this tract
(Tang et al., 2018). All three ROIs are visualized together in MNI152 linear, 1mm resolution
space, in Figure 10.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
46

Figure 10: The LC, Ascending Noradrenergic bundle, and Frontopontine Atlases.

Note Fig. 1. displays the locus coeruleus (yellow), noradrenergic bundle (red), and frontopontine
(blue) tracts registered to MNI152 space. Please note the ascending noradrenergic bundle is a
continuous tract but is not fully displayed here.

Each atlas was registered into MNI152 linear, 1mm brian space before being backwarped
into subject-space using ANTS nonlinear registration, with nearest neighbor interpolation
(Avants et al., 2008). Using BrainSuite (v19b), registrations of each atlas, in each hemisphere,
for each subject, were manually verified by two researchers independently (Shattuck & Leahy,
2002). Confirming accurate atlas registrations for each subject at each timepoint, MRtrix (v3.1)
was used to create a binary mask that was then automatically thresholded. We then computed
fractional anisotropy and axial diffusivity values for the LC, ascending noradrenergic bundle,
and frontopontine masks in each hemisphere. Using R (v4.1.2) and RStudio (v2021.09.2+382),
RUNNING HEAD: AGING AND LC DIFFUSIVITY
47
diffusion values were averaged for each atlas across hemispheres, so that each participant had
one diffusivity value per time point, per atlas (Core Team, 2019; Team, RStudio, 2020).
2.2. Statistical Analyses

Statistical analyses were performed using R and various packages (Bates et al., 2015; Lenth,
2021; Simpson, 2022; Sjoberg et al., 2021; Wickham et al., 2019; Wood, 2011). First, we aimed
to replicate our previous findings of greater fractional anisotropy in the LC of older adults,
relative to younger adults in a cross-sectional comparison. In addition, given the nonlinear
trajectory of diffusivity across the lifespan, generalized additive models (GAMs) were used to
assess diffusivity in each atlas as a function age in the cross-sectional cohort (Lucey et al., 2021;
McLean et al., 2014; Roalf et al., 2020, 2020; Sørensen, Brandmaier, et al., 2021). GAMs
replace the sum of linear terms with the sum of smooth functions and estimate the coefficients
using the data instead of approximating the best polynomial fit (Sørensen, Walhovd, et al.,
2021). This results in penalties across the nonlinear smooth terms to avoid overfitting of the
model (Pedersen et al., 2019; Roalf et al., 2020).
We first assessed the relationship with each ROI and gender. The cross-sectional GAM
equation used for each ROI is shown in equation 1:
Equation 1: Cross-sectional GAM Equation for each ROI
𝐷𝑖𝑓𝑓𝑢𝑠𝑖𝑜𝑛 𝑀𝑒𝑡𝑟𝑖𝑐 ~ 𝑠(𝐴𝑔𝑒) +𝐺𝑒𝑛𝑑𝑒𝑟 + 𝑠(𝐴𝑔𝑒, 𝑏𝑦 = 𝐺𝑒𝑛𝑑𝑒𝑟,𝑏𝑠 = ′𝑓𝑠′),𝑚𝑒𝑡ℎ𝑜𝑑 = "𝑅𝐸𝑀𝐿"
Here, s() indicates a spline fit was applied to the age variable and a smooth term across age, split
by gender. Factor smooth interactions (bs = ‘fs’) are selected to maintain the same smoothing
parameters for each gender. The number or knots selected, specifying the basis function of the
smoothing parameters, were kept at the default. To estimate the coefficients and smoothness
parameters we used the recommended method of restricted maximum likelihood (REML).
RUNNING HEAD: AGING AND LC DIFFUSIVITY
48
In post-hoc analyses, we used marginal derivatives to infer significant relationships
between changes in diffusivity and age in each ROI. Specifically, marginal linear effects of the
smooth term for age are plotted and reported (Makowski et al., 2020). To quantitatively compare
significant differences between smooth terms of each region of interest, we included all three
ROIs into a single model (Equation 2), controlled for gender. Briefly, the smooth term of age for
each ROI is subtracted from other ROIs to calculate a difference in smooth term for age.
Standard errors are then computed to calculate a 95% confidence interval for each smooth pair
difference; more details can be found here (Rose et al., 2012).
Equation 2: GAM Smooth Pair Differences Equation
𝐷𝑖𝑓𝑓𝑢𝑠𝑖𝑜𝑛 𝑀𝑒𝑡𝑟𝑖𝑐 ~ 𝑠(𝐴𝑔𝑒, 𝑏𝑦 = 𝑅𝑂𝐼) + 𝑅𝑂𝐼+ 𝐺𝑒𝑛𝑑𝑒𝑟,𝑚𝑒𝑡ℎ𝑜𝑑 = "𝑅𝐸𝑀𝐿"
Relationships between Rey Auditory Verbal-Learning Test (RAVLT) delay scores and
diffusivity were also analyzed. Because the cross-sectional data set was reduced by over 50
participants when including RAVLT delay scores due to missing values, we analyzed cognitive
associations with separate models (Equation 3). In these models, we placed our subjects into two
groups (young [16-35] and older [> 35]) to compare RAVLT and diffusivity between age groups.

Equation 3: Cross-sectional GAM Equation for RAVLT Delay Scores
𝐷𝑖𝑓𝑓𝑢𝑠𝑖𝑜𝑛 𝑀𝑒𝑡𝑟𝑖𝑐 ~ (𝐴𝑔𝑒𝐺𝑟𝑜𝑢𝑝) +𝐺𝑒𝑛𝑑𝑒𝑟 + 𝑠(𝑅𝐴𝑉𝐿𝑇, 𝑏𝑦 = 𝐴𝑔𝑒𝐺𝑟𝑜𝑢𝑝,𝑏𝑠 = ′𝑓𝑠′),𝑚𝑒𝑡ℎ𝑜𝑑
= "𝑅𝐸𝑀𝐿"
Finally, we removed any outliers that were skewing our models using boxplot quartiles as
criteria for exclusion and to avoid errors in R we scaled axial diffusivity values to center around
1 given their ~1e4 size.
For the longitudinal cohort we adapted General Additive Mixed Models (GAMM) with
the gamm4 package in R utilizing nonlinear mixed modeling (Roalf et al., 2020; Sørensen,
Walhovd, et al., 2021). We used the R syntax described in Sørensen, et al, 2020 section 3.2.1 to
RUNNING HEAD: AGING AND LC DIFFUSIVITY
49
build and check our models for our analyses. Notably, the longitudinal data did not include all
age groups which resulted in an age gap between young and older adults. To circumvent this, we
split the data into young (6 to 18) and older adults (40 to 78), full age range and demographic
breakdown is reported in Table 21. We applied separate GAMM models, one for each age group.
Equation 4 displays the longitudinal cohort model with random intercepts for each subject.

Equation 4: GAMM Equation for the Longitudinal Cohort
𝐷𝑇𝐼 𝑀𝑒𝑡𝑟𝑖𝑐 ~ 𝑠(𝐴𝑔𝑒,𝑏𝑦 = 𝑆𝑒𝑥) +𝑆𝑒𝑥 ,𝑟𝑎𝑛𝑑𝑜𝑚 = ~ (1 | 𝐼𝐷)
Finally, we included all tracts in a single model to compare differences between
longitudinal smooth terms as shown in Equation 5.

Equation 5: GAMM Equation Longitudinal Differences between ROIs
𝐷𝑖𝑓𝑓𝑢𝑠𝑖𝑜𝑛 𝑀𝑒𝑡𝑟𝑖𝑐 ~ 𝑅𝑒𝑔𝑖𝑜𝑛 + 𝑠(𝐴𝑔𝑒,𝑏𝑦 = 𝑅𝑒𝑔𝑖𝑜𝑛) + 𝐺𝑒𝑛𝑑𝑒𝑟,𝑟𝑎𝑛𝑑𝑜𝑚 = ~ (0 + 𝑅𝑒𝑔𝑖𝑜𝑛 | 𝐼𝐷)
Specifically, regions were transformed into ordered factor variables in R to allow the
frontopontine tract to act as a reference intercept in relation to the LC and ascending
noradrenergic bundle. Random intercepts for each region and random slopes per subject were
included in the model. Equation 6 displays the model for RAVLT delay scores within each ROI
over time. All smooth terms p-values were false discovery rate adjusted.

Equation 6: GAMM Equation Longitudinal Differences between RAVLT Delay Scores and
ROIs
𝐷𝑖𝑓𝑓𝑢𝑠𝑖𝑜𝑛 𝑀𝑒𝑡𝑟𝑖𝑐 ~ 𝑅𝑒𝑔𝑖𝑜𝑛 + 𝑠(𝑅𝐴𝑉𝐿𝑇,𝑏𝑦 = 𝑅𝑒𝑔𝑖𝑜𝑛) + 𝐺𝑒𝑛𝑑𝑒𝑟,𝑟𝑎𝑛𝑑𝑜𝑚 = ~ (0 + 𝑅𝑒𝑔𝑖𝑜𝑛 | 𝐼𝐷)

3. Results

3.1. Diffusivity in the LC between Young and Older Adults

RUNNING HEAD: AGING AND LC DIFFUSIVITY
50
We first attempted to replicate previous findings of atypical age effects in the LC
between young and older adults. Since the NKI cohort includes children, we created a temporary
new age criteria from the cross-sectional cohort to match typical data sets of young (ages: 18-30)
and older (ages: 65+) adults, final N = 130. Figure 11 displays our replication of previous
findings that show increased fractional anisotropy in the LC of older adults relative to younger
adults t(130) = -3.681, p < 0.01. However, we did not replicate a significant difference between
young and older adults in fractional anisotropy for the ascending noradrenergic bundle t(130) = -
0.649, p = 0.51.
Figure 11: Fractional Anisotropy Differences between Young and Older Adults in the LC and
the Ascending Noradrenergic Bundle

Note. Figure 11A displays fractional anisotropy in the ascending noradrenergic bundle for young
and older adults. Figure 11B displays fractional anisotropy in the LC for young and older adults.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
51
3.2. Diffusivity Across the Lifespan in the Cross-Sectional Cohort

We display LC, ascending noradrenergic bundle, and frontopontine fractional anisotropy
GAM results in Tables 22A, 22B, 22C and Figures 12A, 12B, and 12C, respectively. Of note,
when a smooth term edf is reported near 0, then it has been removed from the model in a similar
fashion to that of ridge regression shrinking a coefficient near 0. Smooth term edf greater than
one indicate non-linear relationships. In Figures 12A, 12B, and 12C, shaded regions represent
95% confidence intervals. Marginal derivatives are plotted below each corresponding ROI GAM
figure to complement interpretation of GAM models in the LC, ascending noradrenergic bundle,
and frontopontine in Figures 12D, 12E, and 12F, respectively.
In Table 22A and Figure 12A, we report and observe significant nonlinear smooth terms
with fractional anisotropy in the LC peaking around age 60 and declining into late life. Figure
3D displays marginal derivatives of fractional anisotropy and age in the ascending LC. We
observed significant positive association between fractional anisotropy and ages 35-48 (average
marginal effect = 0.004, 95% CI [0.001, 0.007], p = 0.01) and a significant negative association
between fractional anisotropy and ages 66-80, (average marginal effect = -0.002, 95% CI [-
0.004, 0.0002], p = 0.03). Full results are reported in the supplementary section.
In the Ascending Noradrenergic bundle, we observe a peak in fractional anisotropy
around age 60, followed by a steep decline into late life (Table 22B and Figure 12B).
Specifically, we report significant positive relationships in the ascending noradrenergic bundle
fractional anisotropy around ages 9-10 (average marginal effect = 0.0025, 95% CI [6.0098e-
05, 5.0252e-03], p = 0.045) and ages 33-42 (average marginal effect = 0.0027, 95% CI
[6.0765e-04, 4.7949e-03], p = 0.045). Significant negative associations are also observed
between fractional anisotropy at ages 24-25 (average marginal effect = -0.0022, 95% CI [-
RUNNING HEAD: AGING AND LC DIFFUSIVITY
52
4.3944e-03, -4.8325e-05], p = 0.01) and ages 63-85 (average marginal effect = -0.0023, 95% CI
[ -3.7687e-03, -9.0444e-04], p < 0.01), respectively (Figure 12E). Full results are reported in the
supplementary section.
Table 22C and Figure 12C report and display a significant smooth term with age and
fractional anisotropy in the frontopontine tract. There were no significant smooth terms
associated with gender in the frontopontine. Figure 12F displays marginal derivatives for axial
diffusivity and age in the frontopontine tract. We observed significant positive associations
between ages 35-45 (average marginal effect = 0.004, 95% CI [0.0007, 0.007], p = 0.02) and a
significant negative association between ages 65-80, (average marginal effect = -0.003, 95% CI
[-0.005, 0.0008], p = 0.01) (Figure 3F). Full results are reported in the supplementary section.
Lastly, we modeled a smooth term for RAVLT delay scores by age for fractional
anisotropy in each ROI (Equation 3). We observed a significant association between RAVLT
delay scores and fractional anisotropy in the LC, but it did not survive correction (p = 0.061,
FDR corrected) (see supplementary section). Figure 13 displays significant smooth terms for
RAVLT delay scores in older adults (p < 0.01, FDR corrected) and fractional anisotropy in the
ascending noradrenergic bundle; full results are reported in supplementary tables. We did not
observe any significant smooth term of RAVLT delay scores in the frontopontine tract (see
supplementary section).

RUNNING HEAD: AGING AND LC DIFFUSIVITY
53
Table 22: GAM Results for Fractional Anisotropy in the Ascending Noradrenergic Bundle,
Locus Coeruleus, and Frontopontine Tract, Cross-Sectional Cohort
A. Locus Coeruleus
Component Term Estimate Std Error t-value p-value q-value
A. parametric coefficients
(Intercept) 0.369 0.005 72.764 ***
SexFemale -0.006 0.007 -0.861

B. smooth terms
s(Age) 3.413 4.404 8.271 *** < 0.001
s(Age):SexMale 1.000 1.000 4.402 * 0.049
s(Age):SexFemale 1.000 1.000 21.620 *** < 0.001
Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1
q-value = false discovery rate adjusted

Adjusted R-squared: 0.176, Deviance explained 0.193
-REML : -427.911, Scale est: 0.00364, N: 324

B. Ascending Noradrenergic Bundle
Component Term Estimate Std Error t-value p-value q-value
A. parametric coefficients
(Intercept) 0.348 0.002 156.350 ***
SexFemale 0.005 0.003 1.548

B. smooth terms
s(Age) 5.808 7.381 11.547 *** < 0.001
s(Age):SexMale 1.000 1.001 3.141 . 0.2
s(Age):SexFemale 1.000 1.000 0.107

0.7
Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1
q-value = false discovery rate adjusted

Adjusted R-squared: 0.230, Deviance explained 0.252
-REML : -660.804, Scale est: 0.000659, N: 308

C. Frontopontine Tract
Component Term Estimate Std Error t-value p-value q-value
A. parametric coefficients
(Intercept) 0.458 0.004 102.608 ***
SexFemale -0.005 0.006 -0.765

B. smooth terms
s(Age) 3.924 5.067 8.340 ***
s(Age):SexMale 1.000 1.000 1.499

0.3
s(Age):SexFemale 1.000 1.000 2.012

0.3
Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1
q-value = false discovery rate adjusted

Adjusted R-squared: 0.136, Deviance explained 0.155
-REML : -465.754, Scale est: 0.00272, N: 319

RUNNING HEAD: AGING AND LC DIFFUSIVITY
54
Figure 12: GAM Fit and Marginal Derivatives for FA in each Region of Interest

Note. Here we display fractional anisotropy across ages and between genders in the cross-
sectional cohort for the LC, ascending noradrenergic bundle and frontopontine tract in Figures
12A, 12B, and 12C, respectively. Marginal derivatives are also plotted to complement GAM
interpretation for the LC, ascending noradrenergic bundle, and frontopontine in Figures 12D,
12E, and 12F, respectively. Figure 12A displays a cubic fractional anisotropy trajectory across
age in the LC. Figure 12B displays fractional anisotropy changes across age in the ascending
noradrenergic bundle with a steep decline observed around age 60. Figure 12C displays
fractional anisotropy in the frontopontine with a steep decline observed around age 60. Figures
12D, 12E, and 2F display the significant and nonsignificant positive (above zero midline) and
negative (below zero midline) relationships between fractional anisotropy and age in the LC,
ascending noradrenergic bundle, and frontopontine, respectively.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
55
Figure 13: GAM Relationship between Fractional Anisotropy and RAVLT Delay Scores in
Young and Older Adults, Cross-Sectional Cohort

Note: Mean fractional anisotropy in the ascending noradrenergic bundle is positively associated
with RAVLT delay scores in older adults. Young adults in comparison show no relationship. We
did not observe gender differences.

GAM results for axial diffusivity in the LC, ascending noradrenergic bundle, and
frontopontine, are reported in Table 23A, 23B, 23C and displayed in Figures 14A, 14B, and 14C,
respectively. In Figure 14A, 14B, and 14C, shaded regions represent 95% confidence intervals.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
56
Marginal derivatives are plotted below each corresponding ROI to complement interpretation of
GAM models for the LC, ascending noradrenergic bundle, and frontopontine tract in Figures
14D, 14E, and 14F, respectively.
In Table 23A and Figure 14A, we report and observe significant nonlinear smooth terms
with axial diffusivity in the LC, respectively. Unlike the ascending noradrenergic bundle, axial
diffusivity in the LC is observed to decrease into young adulthood, increasing into late life at
which a peak is observed around age 60. Figure 14D complements Figure 14A, displaying
significant negative associations between LC axial diffusivity and ages 6-25 (average marginal
effect = -0.06, 95% CI [-0.10, -0.02], p < 0.01), significant positive associations between ages
35-45, (average marginal effect = 0.08, 95% CI [0.02, 0.13], p = 0.01) and significant negative
associations between ages 67-75, (average marginal effect = -0.09, 95% CI [-0.16, -0.02], p =
0.01). Full results are reported in the supplementary section.
Axial diffusivity increased in the ascending noradrenergic bundle with a significant
nonlinear smooth term increasing from age 40 into late life (Table 23B and Figure 14B).
Specifically, we observed a significant negative associations in the ascending noradrenergic
bundle axial diffusivity between ages 6-26, (average marginal effect = 0.04, 95% CI [-0.07, -
0.01], p < 0.01) and significant positive associations around ages 41 to 85, (average marginal
effect = 0.05, 95% CI [0.02, 0.09], p < 0.01). Full results are reported in supplementary tables.
We report and observe significant nonlinear smooth terms with axial diffusivity in the
frontopontine tract in Table 23C and Figure 14C, respectively. Unlike the other two ROIs, axial
diffusivity in the frontopontine is observed to decrease into young adulthood where it continues
steadily into late life (Figure 14C). Figure 14F displays a significant negative association
between the frontopontine tract and axial diffusivity around ages 6-25 (average marginal effect =
RUNNING HEAD: AGING AND LC DIFFUSIVITY
57
-0.06, 95% CI [-0.10, -0.02], p = 0.01). Full results are reported in supplementary tables. No
significant associations were observed between axial diffusivity and RAVLT delay scores in
young or older adults’ LC, ascending noradrenergic bundle, and frontopontine tract; full results
are reported in the supplementary section.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
58
Table 23: GAM results for Axial Diffusivity in the Ascending Noradrenergic Bundle, LC, and
Frontopontine Tract, Cross-Sectional Cohort
A. Locus Coeruleus
Component Term Estimate Std Error t-value p-value q-value
A. parametric coefficients
(Intercept) -0.105 0.073 -1.442

SexFemale 0.194 0.099 1.968 *
B. smooth terms
s(Age) 5.252 6.715 9.547 *** < 0.001
s(Age):SexMale 2.379 2.959 1.864

.20
s(Age):SexFemale 1.000 1.000 0.264

.60
Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1
q-value = false discovery rate adjusted

Adjusted R-squared: 0.230, Deviance explained 0.253
-REML : 420.640, Scale est: 0.725, N: 324

B. Ascending Noradrenergic Bundle
Component Term Estimate Std Error t-value p-value q-value
A. parametric coefficients
(Intercept) 0.047 0.070 0.679

SexFemale -0.130 0.094 -1.378

B. smooth terms
s(Age) 2.973 3.940 8.846 *** < 0.001
s(Age):SexMale 2.493 3.096 19.656 *** < 0.001
s(Age):SexFemale 1.000 1.000 43.403 *** < 0.001
Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1
q-value = false discovery rate adjusted

Adjusted R-squared: 0.340, Deviance explained 0.356
-REML : 377.323, Scale est: 0.639, N: 308

C. Frontopontine Tract
Component Term Estimate Std Error t-value p-value q-value
A. parametric coefficients
(Intercept) -0.288 0.075 -3.820 ***
SexFemale 0.475 0.102 4.668 ***
B. smooth terms
s(Age) 2.015 2.733 6.602 *** < 0.001
s(Age):SexMale 1.992 2.493 8.857 *** < 0.001
s(Age):SexFemale 1.000 1.000 5.600 * 0.019
Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1
q-value = false discovery rate adjusted

Adjusted R-squared: 0.203, Deviance explained 0.218
-REML : 419.842, Scale est: 0.780, N: 319
Figure 14: GAM Fit and Marginal Derivatives for Axial Diffusivity in each ROI
RUNNING HEAD: AGING AND LC DIFFUSIVITY
59

Note. Here we display axial diffusivity across ages and between genders in the cross-sectional
cohort for the LC, ascending noradrenergic bundle, and frontopontine tract in Figures 14A, 14B,
and 14C, respectively. Marginal derivatives are also plotted to complement GAM interpretation
for the LC, ascending noradrenergic bundle, and frontopontine in Figures 14D, 14E, and 14F,
respectively. Figure 14A displays a nonlinear trajectory of axial diffusivity across age in the LC.
Figure 14B displays axial diffusivity changes across age in the ascending noradrenergic bundle
with a steep incline observed around age 40 through late life. Figure 14C displays axial
diffusivity in the frontopontine with a consistent flat trajectory across ages. Figures 14D, 14E,
and 14F display the significant or nonsignificant positive (above zero midline) and negative
(below zero midline) relationships between fractional anisotropy and age in the LC, ascending
noradrenergic bundle, and frontopontine, respectively.

3.3. Differences between each ROI Smooth Terms for Fractional Anisotropy and Axial
Diffusivity

We subtracted the GAM model smooth term of age between each ROI and plotted with 95%
confidence intervals to identify differences between age-related smooth terms. Anything outside
the zero midline is considered significant. We observed a significant increase in the difference of
fractional anisotropy between the LC and the ascending noradrenergic bundle, Figure 15A.
Figure 15B displays the difference between frontopontine and LC smooth terms with a
significant decline in diffusivity observed in late life. Figure 15C displays the difference in age
RUNNING HEAD: AGING AND LC DIFFUSIVITY
60
smooth terms between the frontopontine and ascending noradrenergic bundle showing a slight
significant increase into midlife with a decrease in late life.
Figure 15: Difference between ROI Smooth Terms across Age in Fractional Anisotropy

Note. Figure 15 displays the difference between GAM model fits for each ROI and fractional
anisotropy across age in the cross-sectional cohort. Shaded regions denote 95% confidence
interval. Figure 15A displays the difference between the LC and ascending noradrenergic bundle.
Figure 15B displays the difference between the frontopontine tract and LC, and Figure 15C
displays the difference between frontopontine tract and the ascending noradrenergic bundle.

Figure 16A displays the difference in smooth terms for age between the LC and
ascending noradrenergic bundle axial diffusivity, with a significant increase into midlife
followed by a significant drop into late life. Figure 16B displays the difference in smooth terms
for age between the frontopontine tract and LC with significant nonlinear differences throughout
early and midlife ages. In Figure 16C, we observe a significant decline in axial diffusivity
differences between the frontopontine and ascending noradrenergic bundle into late life.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
61
Figure 16: Difference between ROI Smooth Terms across Age in Axial Diffusivity

Note. Figure 16 displays the difference between GAM model fits for each ROI and axial
diffusivity across age in the cross-sectional cohort. Figure 16A displays the difference between
the LC and ascending noradrenergic bundle. Figure 16B displays the difference between the
frontopontine tract and LC, and Figure 16C displays the difference between frontopontine tract
and the ascending noradrenergic bundle.

3.4. Longitudinal Trajectories in Diffusivity in Young and Older Adults

Longitudinal GAMM age smooth term plots for fractional anisotropy of the LC, ascending
noradrenergic bundle, and frontopontine tract are displayed in Figure 17A, 17B, 17C,
respectively. We did not observe a significant smooth term for age in either young (male: p >
0.9, female: p = 0.2, FDR corrected) or older adults (male: p = 0.3, female: p = 0.4, FDR
corrected) in the LC (Figures 17A and 17D, respectively), see supplementary tables. We
identified significant smooth terms for age by gender in young adults (both male and female, p <
0.01 FDR corrected) but not older adults (male: p = 0.2, female: p = 0.7, FDR corrected) in the
ascending noradrenergic bundle, Figures 17B and 17E and supplementary tables , respectively.
In the frontopontine tract, we only observed significant terms for age in young adult males (p =
0.003, FDR corrected) but not females (p = 0.2, FDR corrected). We did not observe any
RUNNING HEAD: AGING AND LC DIFFUSIVITY
62
significant smooth term for age in older adults (male: p = 0.3, female: p = 0.6, FDR corrected), in
the frontopontine tract, Figures 17C and 17F, respectively and supplementary tables.

Figure 17: Longitudinal Changes in Fractional Anisotropy by Age and Gender

Note: Fractional anisotropy estimated over age (young or older adults) between genders (males
or females) in the ascending noradrenergic bundle (Figure 17A and 17D), the LC (Figure 17B,
and 17E) and the frontopontine (Figure 17C, and 17F) tract. Raw data with a line connecting
each subject’s visit (baseline, follow up 1, and follow up 2) is displayed in addition to the
GAMM estimates.

Longitudinal GAMM age smooth term plots for axial diffusivity in the LC, ascending
noradrenergic bundle, and frontopontine tract are displayed in Figure 18A, 18B, 18C,
respectively. We did not observe a significant smooth term for age in either young (male: p =
0.3, female: p = 0.4, FDR corrected) or older adults (male: p = 0.3, female: p = 0.3, FDR
corrected) in the LC, Figures 18A and 18D, respectively and supplementary tables, respectively.
We observed significant smooth terms for age by gender in young adult males and trending in
females (male: p = 0.005, female: p = 0.71, FDR corrected), as well as significant terms in older
adults (male: p < 0.001, female: p < 0.001, FDR corrected) in the ascending noradrenergic
bundle, Figures 18B and 18E, and supplementary tables, respectively. In the frontopontine tract,
RUNNING HEAD: AGING AND LC DIFFUSIVITY
63
we only observed significant terms for age in young adults (male: p = 0.025, females: (p = 0.18,
FDR corrected). We did not observe any significant smooth term for age in older adults (male: p
= 0.6, female: p = 0.6, FDR corrected) in the frontopontine, Figures 18C and 18F, and
supplementary tables, respectively.

Figure 18: Longitudinal Changes in Axial Diffusivity by Age and Gender

Note. Axial Diffusivity estimated over age (young or older adults) between genders (males or
females) in the ascending noradrenergic bundle (Figure 18A and 18D), the LC (Figure 18B, and
18E) and the frontopontine (Figure 18C, and 18F) tract. Raw data with a line connecting each
subject’s visit (baseline, follow up 1, and follow up 2) is displayed below the GAMM estimates.

3.5. Comparison of Longitudinal Trajectories between Regions of Interest

In our final analyses, we included all ROIs into a single model and set the frontopontine tract
as an intercept which allowed us to compare the frontopontine age smooth term to the LC and
ascending noradrenergic bundle age smooth terms. Full results reporting fixed, random and
GAM effects are available in supplementary tables. Supplementary Figure 19 displays the 95%
confidence interval for the fixed effects portion of the model. In the young adult fractional
anisotropy model, there was no significant smooth term for the LC, relative to the frontopontine
RUNNING HEAD: AGING AND LC DIFFUSIVITY
64
intercept (p = 0.2, FDR corrected), while a significant smooth term of age for the ascending
noradrenergic bundle was observed (p < 0.01, FDR corrected). In the older adult fractional
anisotropy model, we observed a trend in the LC smooth term of age (p = 0.069, FDR corrected),
but not in the ascending noradrenergic bundle (p = 0.2, FDR corrected).
For axial diffusivity, the LC smooth term for young adults was significant (p < 0.01, FDR
corrected), but trending in older adults (p = 0.056, FDR corrected). In the ascending
noradrenergic bundle, young adults displayed a trending smooth term relative to the
frontopontine tract (p = 0.072, FDR corrected) while older adults displayed a significant smooth
term for age (p < 0.001). We observed no significant interaction between ROIs and RAVLT
delay scores across time, results reported in supplementary Table 50 and 51, and 95% confidence
interval for GAMM fixed terms in supplementary Figure 20.
4. Discussion

We investigated differences in diffusivity across the lifespan of the LC and its ascending
noradrenergic bundle. The frontopontine tract was included as a control region, as it travels
ventrally to the LC (Zhang et al., 2017). We successfully replicated the atypical age-related
findings of older adults displaying higher fractional anisotropy in the LC, in comparison with
younger adults (Figure 11B) (Langley et al., 2020; Porat et al., 2022). However, we were not
successful in replicating age-expected differences in fractional anisotropy of the ascending
noradrenergic bundle (Figure 11A). Generalized additive models were chosen for their flexibility
to apply nonlinear smooth terms as a more robust approach to measure diffusivity changes across
age groups (Lucey et al., 2021; Muncy et al., 2022; Sørensen, Brandmaier, et al., 2021; Winter &
Wieling, 2016). A significant smooth term of age was identified in three ROIs and each ROI
displayed unique trajectories across the lifespan. Specifically, expected age-related declines in
RUNNING HEAD: AGING AND LC DIFFUSIVITY
65
fractional anisotropy were observed in the ascending noradrenergic bundle (Figure 12B), with
less pronounced changes observed in the frontopontine and LC (Figure 12C and 12A).
Similarly, expected increases in axial diffusivity were observed with a steeper slope into
late life in the ascending noradrenergic bundle (Figure 13B), but this was not the case in the LC
or frontopontine (Figure 13A and 13C). We partially replicated previous reports of associations
with RAVLT delay scores and LC and axial diffusivity in older adults, however these were not
significant after p-value adjustments, (see supplementary tables) (Langley et al., 2020). Other
studies have also reported LC and memory relationships with non-diffusion metrics like LC-MRI
contrast ratios (Clewett et al., 2018; Dahl et al., 2019, 2022). RAVLT has been reported to
predict poor cognitive performance and dementia onset, so it is possible that our healthy cohort
may be the reason for the lack of significant findings (Dahl et al., 2022; Moradi et al., 2016).
Interestingly, fractional anisotropy in the LC declined between early childhood until the
mid-30s, where thereafter a gradual incline is observed until a peak around 60 years of age
(Figure 12B). This begs the question of what inferences can be deduced from a typical young vs
older comparison of LC fractional anisotropy? Because there were no associations between
young adults’ LC diffusivity and RAVLT, we cannot assume that less fractional anisotropy in
young adults is detrimental to memory, though further investigation is necessary. Beyond age 60,
fractional anisotropy in the LC was observed to reduce gradually, suggesting the LC may still
follow natural age-related declines in diffusion but much later in life. However, in comparison to
the ascending noradrenergic bundle, fractional anisotropy of the LC increased into late life.
Given the non-linear nature of fractional anisotropy and axial diffusion observed in the LC and
ascending noradrenergic bundle throughout the lifespan, future studies should explore underlying
mechanisms that may be modulating diffusivity. Additionally, biomarkers aimed to identify age-
RUNNING HEAD: AGING AND LC DIFFUSIVITY
66
related changes in diffusivity may be unique in the LC in relation to other regions of interest.
High resolution diffusion MRI and diffusion techniques like NODDI may aim to replicate our
findings and expand support to the hypothesis that the LC exhibits unique microstructural
properties. Arousal or attentional tasks may be studied with diffusion MRI to determine
associations with decreased LC fractional anisotropy in young adults or increased diffusivity in
older adults.
We observed significant differences between young and older adult fractional anisotropy
and diffusivity in each region of interest. Specifically, we observed age expected changes in axial
diffusivity and fractional anisotropy in the ascending noradrenergic bundle (Figure 13B).
However, older adults did not exhibit increases in axial diffusivity or reductions in fractional
anisotropy across timepoints within the LC (Figure 13A). In the frontopontine tract, young adults
displayed typical age-expected declines in axial diffusivity and increases in fractional anisotropy,
in comparison to diffusivity within older adults (Figure 13C).
Notably, the cubic trajectory of LC fractional anisotropy and axial diffusivity could be
speculated to associate with functional connectivity of the LC. Although no studies have directly
investigated functional connectivity in the LC in relation to diffusivity, two studies have recently
reported cubic and quadratic relationships with functional connectivity in the LC (H. I. L. Jacobs
et al., 2018; Song et al., 2021). The quadratic association was observed in the same NKI data set
with a decrease in connectivity with sensory regions into midlife (similar to a decrease in
fractional anisotropy as shown in Figure 12A) but the inverse was observed in frontal regions
(Song et al., 2021). Fractional anisotropy has been positively correlated to functional
connectivity, though given the LC’s gray matter anatomy, future studies should explore these
associations further (Babaeeghazvini et al., 2021).
RUNNING HEAD: AGING AND LC DIFFUSIVITY
67
Recent studies have also investigated the relationship between diffusivity and age across
the lifespan (Beck et al., 2021, 2021; Kochunov et al., 2012). A nonlinear negative relationship
between fractional anisotropy and positive relationship with axial diffusivity is reported between
ages 20 and 85 (Beck et al., 2021). Additionally, the ascending noradrenergic bundle in the
longitudinal cohort displayed significant positive linear associations of axial diffusivity with
older adults’ age, similar to studies that only explored aging cohorts (Merenstein et al., 2021).
However, the LC and frontopontine did not adhere to the same degrees of decline in diffusivity
with increasing age. Nonlinear trajectories of diffusion are important to study because they help
us determine the potential ages of normal degeneration versus that of pathological diseases like
Alzheimer’s (Beck et al., 2021; Merenstein et al., 2021).
Many studies enroll participants aged at least 18 years old for younger adults, but when
investigating diffusivity changes in early childhood and adolescence, studies have reported
increases in fractional anisotropy and decreases in axial and mean diffusivity (Cascio et al.,
2013; Kochunov et al., 2012; Westlye et al., 2010). We observed such early childhood fractional
anisotropy trends in the ascending noradrenergic bundle but not in the LC or frontopontine tract.
But axial diffusivity did decrease in early life within all three ROIs (Figure 12). Together, these
findings also raise the question of why the LC is observed to have different diffusivity
trajectories across the lifespan.
The first aspect that comes to mind is that the LC is a gray matter deep brain nucleus
(Mather & Harley, 2016) whereas DTI is typically measuring white matter microstructure
(Soares et al., 2013). However, several studies have reported decreased fractional anisotropy and
increased diffusivity in gray matter regions to be associated with aging (Henf et al., 2018; Ni et
al., 2010; Salminen et al., 2016). One study investigated deep gray matter structures and reported
RUNNING HEAD: AGING AND LC DIFFUSIVITY
68
that fractional anisotropy (as well as diffusivity) was higher in older adults, relative to younger
adults, potentially due to increased iron deposits (Pfefferbaum et al., 2010). Could this explain
our findings? The answer is unclear as iron is more abundant in the substantia nigra than LC
(Langley et al., 2020; Zucca et al., 2006). And previous studies suggest that iron deposition
cannot be associated with LC diffusivity due to the LC’s neuromelanin bond with copper (Ju et
al., 2013; Langley et al., 2020; Sulzer et al., 2018). This may not be the case with frontopontine
tracts that project fibers through the substantia nigra and may be exposed to iron deposits at a
subcortical level (Zhang et al., 2017).
Tau pathology has been associated with decreased fractional anisotropy in white matter
tracts as well (Sintini et al., 2018; Strain et al., 2018). Given the role of the LC in producing
abnormal tau hyperphosphorylation, could LC activity cause downstream diffusivity changes
that are observed in the ascending noradrenergic bundle (Chalermpalanupap et al., 2017)?
Several studies in the past year have identified a negative relationship between tau and the
ascending noradrenergic bundle diffusivity (Chu et al., 2022; Quattrini et al., 2021; Solders et al.,
2022). To our knowledge, no studies have exclusively investigated the relationship between tau
and diffusivity in the LC. One possible underlying mechanism that has yet to be explored and
may explain a reduction in fractional anisotropy in young adults is sleep deprivation.
Recent investigations have observed significant reductions in fractional anisotropy in
young adults following a night of sleep deprivation (Elvsåshagen et al., 2015; Grumbach et al.,
2020; Rocklage et al., 2009; Voldsbekk et al., 2021). Given that young adults are more likely to
be sleep deprived and simultaneously more affected by acute sleep deprivation (Halassa et al.,
2009; Zitting et al., 2018), an association with lower fractional anisotropy may be present.
Specifically, the role of the LC in arousal and sleep is well documented (Hayat et al., 2020;
RUNNING HEAD: AGING AND LC DIFFUSIVITY
69
Khanday et al., 2016; Takahashi et al., 2010). Sleep deprivation is associated with increased
inflammation and depression (Dinis & Bragança, 2018; Dzierzewski et al., 2020; Mullington et
al., 2010; Vanek et al., 2020), both of which are correlated with decreases in fractional
anisotropy and commonly exhibited in young adults (Bijanki et al., 2015; Kulesh et al., 2018;
Meinert et al., 2019; Walker et al., 2017). Further investigations in the associations between
sleep deprivation and fractional anisotropy are needed.
Here, we identified unique nonlinear associations between fractional anisotropy, axial
diffusivity, and age smooth terms in the LC, ascending noradrenergic bundle, and frontopontine
tract. Our results suggest that LC diffusivity does not adhere to typical age-related trajectories
observed in other regions of the brain like the ascending noradrenergic bundle. Differences in
smooth terms of age between the LC and noradrenergic bundle suggest that the LC is less
susceptible to natural age-related degenerations in diffusion and fractional anisotropy.
4.1. Limitations

There are some limitations in our study involving generalized additive modeling. The
difficulty interpreting GAM forces us to rely on post-hoc analyses like marginal derivatives.
Cohorts were not measured at the study site during the same time and longitudinal data was not
always collected at the same time intervals, which may result in collinearity effects.
Additionally, structural equation modeling may provide more information on inter-subject
variability in the longitudinal cohort. The diffusion imaging scans also pose limitations at 2mm
3

resolution where magnetization transfer sequences to image the LC yield better anatomical
resolutions. The cohort in this study was also composed of healthy young and older adults that
possessed no neurological or physiological health issues and may not represent the general aging
population.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
70
5. Supplementary Material

Table 24: Marginal Derivatives for Fractional Anisotropy and Age in the Locus Coeruleus,
Cross-Sectional Cohort
Age
Start
Age
End
Mean
Age
Coefficient SE CI CI
low
CI
high
t df p
6.00 34.71 20.35 0.00 0.00 0.95 0.00 0.00 -0.76 316.59 0.28
34.78 47.99 41.39 0.00 0.00 0.95 0.00 0.01 2.57 316.59 0.01
48.07 66.26 57.16 0.00 0.00 0.95 0.00 0.00 0.00 316.59 0.37
66.34 80.41 73.38 0.00 0.00 0.95 0.00 0.00 -2.20 316.59 0.03
Note. SE = standard error, CI = confidence interval, t = test statistic, df = degrees of freedom.

Table 25: Marginal Derivatives for Fractional Anisotropy and Age in the Ascending
Noradrenergic Bundle, Cross-Sectional Cohort
Age
Start
Age
End
Mean
Age
Coefficient SE CI CI
low
CI
high
t df p
6.00 9.32 7.66 0.00 0.00 0.95 0.00 0.01 1.63 298.19 0.11
9.40 10.43 9.91 0.00 0.00 0.95 0.00 0.01 2.02 298.19 0.04
10.51 24.27 17.39 0.00 0.00 0.95 0.00 0.00 -0.67 298.19 0.28
24.35 25.37 24.86 0.00 0.00 0.95 0.00 0.00 -2.01 298.19 0.05
25.45 32.89 29.17 0.00 0.00 0.95 0.00 0.00 -0.19 298.19 0.34
32.97 41.74 37.35 0.00 0.00 0.95 0.00 0.00 2.54 298.19 0.02
41.82 63.25 52.54 0.00 0.00 0.95 0.00 0.00 0.23 298.19 0.48
63.33 85.00 74.17 0.00 0.00 0.95 0.00 0.00 -3.34 298.19 0.01
Note. SE = standard error, CI = confidence interval, t = test statistic, df = degrees of freedom.

Table 26: Marginal Derivatives for Fractional Anisotropy and Age in the Frontopontine Tract,
Cross-Sectional Cohort
Age
Start
Age
End
Mean
Age
Coefficient SE CI CI
low
CI
high
t df p
6.00 34.63 20.31 0.00 0.00 0.95 0.00 0.00 -0.49 311.08 0.44
34.71 44.99 39.85 0.00 0.00 0.95 0.00 0.01 2.40 311.08 0.02
45.07 64.52 54.79 0.00 0.00 0.95 0.00 0.00 0.18 311.08 0.39
64.60 80.33 72.47 0.00 0.00 0.95 -0.01 0.00 -2.70 311.08 0.01
80.41 85.00 82.71 0.00 0.00 0.95 -0.01 0.00 -1.73 311.08 0.09
Note. SE = standard error, CI = confidence interval, t = test statistic, df = degrees of freedom.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
71

Table 27: Fractional Anisotropy in the Locus Coeruleus with RAVLT, Cross-Sectional Cohort

Table 28: Fractional Anisotropy in the Ascending Noradrenergic Bundle with RAVLT Delay
Scores, Cross-Sectional Cohort
Component Term Estimate
Std
Error
t-value
p-
value

A. parametric
coefficients
(Intercept) 0.354 0.002 159.041 ***
Older Adults 0.007 0.003 2.463 *

SexFemale 0.004 0.003 1.429

Component Term edf Ref. df F-value
p-
value
q-
value
B. smooth terms
s(RAVLT
DELAY):Young
1.112 1.216 0.134
0.9
s(RAVLT
DELAY):Older
1.001 1.002 17.760 ***
<
0.001
Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1 q-value =
False discover rate multiple corrections.

Adjusted R-squared: 0.0802, Deviance explained 0.0947
-REML : -620.614, Scale est: 0.000419, N: 261

Table 29: Fractional Anisotropy in the Frontopontine with RAVLT, Cross-Sectional Cohort
Component Term Estimate Std Error t-value p-value
A. parametric coefficients
(Intercept) 0.460 0.005 96.506 ***
Older Adults 0.025 0.006 3.948 ***
SexFemale -0.003 0.006 -0.569

Component Term edf Ref. df F-value p-value
B. smooth terms
s(RAVLT DELAY):Young 1.001 1.002 0.020

s(RAVLT DELAY):Older 1.435 1.753 3.861

Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1
Adjusted R-squared: 0.0538, Deviance explained 0.0697
Component Term Estimate Std Error t-value p-value q-value
A. parametric coefficients
(Intercept) 0.365 0.006 65.917 ***
Older Adults 0.049 0.007 6.740 ***
SexFemale -0.013 0.007 -1.966 .
Component Term edf Ref.df
F-
value
p-value
q-value
B. smooth terms
s(RAVLT DELAY):Young 1.000 1.000 4.736 * 0.061
s(RAVLT DELAY):Older 1.000 1.000 1.365

Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1, q-value = False
discover rate multiple corrections.

Adjusted R-squared: 0.138, Deviance explained 0.152
RUNNING HEAD: AGING AND LC DIFFUSIVITY
72
Component Term Estimate Std Error t-value p-value
-REML : -424.266, Scale est: 0.00204, N: 265

Table 30: Marginal Derivatives for Axial Diffusivity and Age in the Locus Coeruleus, Cross-
Sectional Cohort
Age
Start
Age
End
Mean
Age
Coefficient SE CI CI
low
CI
high
t df p
6.00 25.37 15.69 -0.06 0.02 0.95
-
0.10
-
0.02
-
3.43 313.37 0.00
25.45 35.26 30.36 0.00 0.02 0.95
-
0.05 0.05
-
0.19 313.37 0.36
35.34 45.22 40.28 0.08 0.03 0.95 0.02 0.13 2.61 313.37 0.01
45.30 67.21 56.25 0.01 0.03 0.95
-
0.05 0.07 0.47 313.37 0.34
67.29 74.88 71.08 -0.09 0.03 0.95
-
0.16
-
0.02
-
2.53 313.37 0.01
74.96 85.00 79.98 0.00 0.05 0.95
-
0.11 0.11
-
0.11 313.37 0.62

Table 31: Marginal Derivatives for Axial Diffusivity and Age in the Ascending Noradrenergic
Bundle, Cross-Sectional Cohort
Age
Start
Age
End
Mean
Age
Coefficient SE CI CI
low
CI
high
t df p
6 26.09 16.04 -0.04 0.01 0.95 -0.07 -0.01 -
2.85
299.53 0.01
26.17 41.74 33.95 0 0.01 0.95 -0.03 0.03 0.08 299.53 0.39
41.82 85 63.41 0.05 0.02 0.95 0.02 0.09 3.08 299.53 0

Table 32: Marginal Derivatives for Axial Diffusivity and Age in the Frontopontine Tract, Cross-
Sectional Cohort
Age
Start
Age
End
Mean
Age
Coefficient SE CI CI
low
CI
high
t df p
6.00 25.69 15.85 -0.06 0.02 0.95 -0.10 -0.02 -3.13 311.99 0.01
25.77 85.00 55.38 0.00 0.01 0.95 -0.02 0.02 0.28 311.99 0.47

RUNNING HEAD: AGING AND LC DIFFUSIVITY
73
Table 33: Axial Diffusivity in the Locus Coeruleus with RAVLT, Cross-Sectional Cohort
Component Term Estimate Std Error t-value p-value
A. parametric coefficients
(Intercept) 0.002 0.000 111.901 ***
Older Adults 0.000 0.000 1.679 .
SexFemale 0.000 0.000 2.169 *
Component Term edf Ref. df F-value p-value
B. smooth terms
s(RAVLT DELAY):Young 1.003 1.006 1.802

s(RAVLT DELAY):Older 1.332 1.594 0.825

Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1
Adjusted R-squared: 0.0408, Deviance explained 0.0566
-REML : -1930.328, Scale est: 1.8e-08, N: 264

Table 34: Axial diffusivity in the Ascending Noradrenergic Bundle with RAVLT Delay Scores,
Cross-Sectional Cohort
Component Term Estimate Std Error t-value p-value
A. parametric coefficients
(Intercept) -0.198 0.097 -2.048 *
Older Adults 0.450 0.124 3.641 ***
SexFemale -0.208 0.115 -1.812 .
Component Term edf Ref. df F-value p-value
B. smooth terms
s(RAVLT DELAY):Young 1.000 1.000 0.021

s(RAVLT DELAY):Older 1.001 1.002 1.027

Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1
Adjusted R-squared: 0.060, Deviance explained 0.0745
-REML : 345.047, Scale est: 0.792, N: 261

Table 35: Axial Diffusivity in the Frontopontine with RAVLT, Cross-Sectional Cohort
Component Term Estimate Std Error t-value p-value
A. parametric coefficients
(Intercept) -0.052 0.097 -0.540

Older Adults -0.538 0.127 -4.246 ***
SexFemale 0.459 0.118 3.891 ***
Component Term edf Ref. df F-value p-value
B. smooth terms
s(RAVLT DELAY):Young 1.000 1.000 0.468

s(RAVLT DELAY):Older 1.000 1.001 1.200

Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1
Adjusted R-squared: 0.0538, Deviance explained 0.0697
-REML : -424.266, Scale est: 0.00204, N: 265

RUNNING HEAD: AGING AND LC DIFFUSIVITY
74
Table 36: GAMM Results for Fractional Anisotropy in the LC, Young Adults – Longitudinal
Cohort
Characteristic exp(Beta) 95% CI
1
p-value q-value
2

Random Effects
(Intercept) 0.69 0.58, 0.82

Female 0.94 0.70, 1.27

s(Age):Male (Fixed) 0.98 0.84, 1.15

s(Age):Female (Fixed) 1.10 0.88, 1.37

Variances
Subject (Intercept)
0.25

s(Age):Female 0.00

s(Age):Male 0.00

Residual 0.38

GAM Portion
Sex

Female -0.06 -0.36, 0.24 0.7 0.8
s(Age):SexMale

0.8 0.8
s(Age):SexFemale

0.4 0.8
1
CI = Confidence Interval,
2
False discovery rate correction for multiple testing

Table 37: GAMM Results for Fractional Anisotropy in LC, Older Adults – Longitudinal Cohort
Characteristic exp(Beta) 95% CI
1
p-value q-value
2

Random Effects
X(Intercept) 1.38 1.01, 1.87

(Intercept) 1.02 0.70, 1.48

Female 1.06 0.62, 1.84

s(Age):Male (Fixed) 1.05 0.44, 2.51

s(Age):Female (Fixed) 0.54

Variances
Subject (Intercept)
1.28

s(Age):Female 0.37

s(Age):Male 0.50

GAM Portion
Sex

Female 0.02 -0.36, 0.39 >0.9 >0.9
s(Age):SexMale

0.8 >0.9
s(Age):SexFemale

0.073 0.2
1
CI = Confidence Interval,
2
False discovery rate correction for multiple testing

RUNNING HEAD: AGING AND LC DIFFUSIVITY
75
Table 38: GAMM Results for Fractional Anisotropy in the Ascending Noradrenergic Bundle,
Young Adults – Longitudinal Cohort
Characteristic exp(Beta) 95% CI
1
p-value q-value
2

Random Effects
(Intercept) 0.73 0.55, 0.96

Female 1.33 0.85, 2.10

s(Age):Male (Fixed) 2.15 0.72, 6.40

s(Age):Female (Fixed) 1.98 1.22, 3.20

Variances
Subject (Intercept)
0.66

s(Age):Female 0.22

s(Age):Male 3.31

Residual 0.67

GAM Portion
Sex

Female 0.29 -0.17, 0.74 0.2 0.2
s(Age):Male

<0.001 <0.001
s(Age):Female

<0.001 <0.001
1
CI = Confidence Interval,
2
False discovery rate correction for multiple testing

Table 39: GAMM Results for Fractional Anisotropy in the Ascending Noradrenergic Bundle,
Older Adults – Longitudinal Cohort
Characteristic exp(Beta) 95% CI
1
p-value q-value
2

Random Effects
(Intercept) 1.03 0.80, 1.33

Female 1.28 0.94, 1.75

s(Age):Male (Fixed) 0.79 0.50, 1.26

s(Age):Female (Fixed) 1.03 0.75, 1.42

Variances
Subject (Intercept)
0.44

s(Age):Female 0.11

s(Age):Male 0.28

Residual 0.15

GAM Portion
Sex

Female 0.25 -0.06, 0.56 0.12 0.2
s(Age):Male

0.074 0.2
s(Age):Female

0.7 0.7
1
CI = Confidence Interval,
2
False discovery rate correction for multiple testing

RUNNING HEAD: AGING AND LC DIFFUSIVITY
76
Table 40: GAMM Results for Fractional Anisotropy in the Frontopontine, Young Adults –
Longitudinal Cohort
Characteristic exp(Beta) 95% CI
1
p-value q-value
2

Random Effects
(Intercept) 0.68 0.53, 0.86

Female 1.07 0.72, 1.60

s(Age):Male (Fixed) 1.35 1.13, 1.60

s(Age):Female (Fixed) 1.28 0.81, 2.02

Variances
Subject (Intercept)
0.61

s(Age):Female 0.27

s(Age):Male 0.00

Residual 0.26

GAM Portion
Sex

Female 0.07 -0.33, 0.47 0.7 0.7
s(Age):SexMale

<0.001 0.003
s(Age):SexFemale

0.11 0.2
1
CI = Confidence Interval,
2
False discovery rate correction for multiple testing

Table 41: GAMM Results for Fractional Anisotropy in the Frontopontine, Older Adults –
Longitudinal Cohort
Characteristic exp(Beta) 95% CI
1
p-value q-value
2

Random Effects
(Intercept) 1.59 1.13, 2.23

Female 0.75 0.50, 1.13

s(Age):Male (Fixed) 1.12 0.49, 2.52

s(Age):Female (Fixed) 1.05 0.85, 1.30

Variances
Subject (Intercept)
0.83

s(Age):Female 0.00

s(Age):Male 1.29

Residual 0.09

GAM Portion
Sex

Female -0.29 -0.70, 0.12 0.2 0.3
s(Age):SexMale

0.2 0.3
s(Age):SexFemale

0.6 0.6
1
CI = Confidence Interval,
2
False discovery rate correction for multiple testing

RUNNING HEAD: AGING AND LC DIFFUSIVITY
77
Table 42: GAMM Results for Axial Diffusivity in the LC, Young Adults – Longitudinal Cohort
Characteristic exp(Beta) 95% CI
1
p-value q-value
2

Random Effects
(Intercept) 0.90 0.71, 1.15

Female 1.15 0.77, 1.72

s(Age):Male (Fixed) 0.84 0.69, 1.03

s(Age):Female (Fixed) 1.00 0.45, 2.24

Variances
Subject (Intercept)
0.51

s(Age):Female 1.11

s(Age):Male 0.00

Residual 0.50

GAM Portion
Sex

Female 0.14 -0.26, 0.54 0.5 0.5
s(Age):SexMale

0.092 0.3
s(Age):SexFemale

0.3 0.4
1
CI = Confidence Interval,
2
False discovery rate correction for multiple testing

Table 43: GAMM Results for Axial Diffusivity in LC, Older Adults – Longitudinal Cohort
Characteristic exp(Beta) 95% CI
1
p-value q-value
2

Random Effects
X(Intercept) 0.74 0.55, 0.99

(Intercept) 1.66 1.16, 2.36

Female 1.20 0.78, 1.84

s(Age):Male (Fixed) 1.12 0.92, 1.38

s(Age):Female (Fixed) 0.51

Variances
Subject (Intercept)
0.00

s(Age):Female 0.19

s(Age):Male 0.39

GAM Portion
Sex

Female 0.51 0.15, 0.86 0.005 0.016
s(Age):SexMale

0.2 0.3
s(Age):SexFemale

0.3 0.3
1
CI = Confidence Interval,
2
False discovery rate correction for multiple testing

RUNNING HEAD: AGING AND LC DIFFUSIVITY
78
Table 44: GAMM Results for Axial Diffusivity in the Ascending Noradrenergic Bundle, Young
Adults – Longitudinal Cohort
Characteristic exp(Beta) 95% CI
1
p-value q-value
2

Random Effects
(Intercept) 1.01 0.83, 1.23

Female 0.78 0.57, 1.08

s(Age):Male (Fixed) 0.77 0.66, 0.90

s(Age):Female (Fixed) 0.80 0.64, 1.00

Variances
Subject (Intercept)
0.34

s(Age):Female 0.00

s(Age):Male 0.00

Residual 0.32

GAM Portion
Sex

Female -0.24 -0.56, 0.08 0.14 0.14
s(Age):SexMale

0.002 0.005
s(Age):SexFemale

0.048 0.071
1
CI = Confidence Interval,
2
False discovery rate correction for multiple testing

Table 45: GAMM Results for Axial Diffusivity in the Ascending Noradrenergic Bundle, Older
Adults – Longitudinal Cohort
Characteristic exp(Beta) 95% CI
1
p-value q-value
2

Random Effects
(Intercept) 0.99 0.73, 1.35

Female 1.09 0.75, 1.59

s(Age):Male (Fixed) 2.13 1.14, 3.97

s(Age):Female (Fixed) 1.75 1.42, 2.16

Variances
Subject (Intercept)
0.65

s(Age):Female 0.00

s(Age):Male 0.56

Residual 0.21

GAM Portion
Sex

Female 0.09 -0.29, 0.46 0.7 0.7
s(Age):SexMale

<0.001 <0.001
s(Age):SexFemale

<0.001 <0.001
1
CI = Confidence Interval,
2
False discovery rate correction for multiple testing

RUNNING HEAD: AGING AND LC DIFFUSIVITY
79
Table 46: GAMM Results for Axial Diffusivity in the Frontopontine, Young Adults –
Longitudinal Cohort
Characteristic exp(Beta) 95% CI
1
p-value q-value
2

Random Effects
(Intercept) 1.34 1.07, 1.67

Female 1.81 1.26, 2.60

s(Age):Male (Fixed) 1.05 0.62, 1.79

s(Age):Female (Fixed) 1.33 1.07, 1.65

Variances
Subject (Intercept)
0.50

s(Age):Female 0.00

s(Age):Male 0.65

Residual 0.23

GAM Portion
Sex

Female 0.59 0.23, 1.0 0.002 0.005
s(Age):SexMale

0.025 0.025
s(Age):SexFemale

0.012 0.018
1
CI = Confidence Interval,
2
False discovery rate correction for multiple testing

Table 47: GAMM Results for Axial Diffusivity in the Frontopontine, Older Adults –
Longitudinal Cohort
Characteristic exp(Beta) 95% CI
1
p-value q-value
2

Random Effects
(Intercept) 0.42 0.31, 0.56

Female 2.10 1.46, 3.00

s(Age):Male (Fixed) 1.11 0.85, 1.44

s(Age):Female (Fixed) 1.05 0.86, 1.27

Variances
Subject (Intercept)
0.63

s(Age):Female 0.00

s(Age):Male 0.00

Residual 0.11

GAM Portion
Sex

Female 0.74 0.38, 1.1 <0.001 <0.001
s(Age):SexMale

0.4 0.6
s(Age):SexFemale

0.6 0.6
1
CI = Confidence Interval,
2
False discovery rate correction for multiple testing

RUNNING HEAD: AGING AND LC DIFFUSIVITY
80
Table 48: Linear Mixed Effects Portion of Longitudinal GAMM Results

Young FA Older FA Young AD Older AD
Fixed Effects
(Intercept) -0.232 0.124 0.031 -0.043

s.e. = 0.045 s.e. = 0.051 s.e. = 0.024 s.e. = 0.024

t = -5.151 t = 2.411 t = 1.268 t = -1.803
FPT Linear -1.135 -1.345 -0.001 0.163

s.e. = 0.039 s.e. = 0.044 s.e. = 0.018 s.e. = 0.017

t = -28.783 t = -30.869 t = -0.067 t = 9.829
FPT Quadratic 0.518 0.225 -1.590 -1.699

s.e. = 0.061 s.e. = 0.064 s.e. = 0.034 s.e. = 0.026

t = 8.533 t = 3.539 t = -47.233 t = -66.423
Gender: Female 0.026 0.074 -0.004 0.029

s.e. = 0.067 s.e. = 0.052 s.e. = 0.025 s.e. = 0.023

t = 0.387 t = 1.435 t = -0.160 t = 1.261
s(Age) × LC 0.077 0.126 -0.142 0.057

s.e. = 0.051 s.e. = 0.308 s.e. = 0.250 s.e. = 0.110

t = 1.528 t = 0.408 t = -0.568 t = 0.516
s(Age) × ANB 0.222 -0.035 -0.024 0.083

s.e. = 0.152 s.e. = 0.062 s.e. = 0.012 s.e. = 0.011

t = 1.466 t = -0.569 t = -1.900 t = 7.255
Random Effects
FPT SD 0.507 0.581 0.186 0.197
FPT x LC Cor -0.018 0.143 0.596 0.473
FPT x ANB Cor 0.925 0.924 0.984 0.382
LC SD 0.492 0.689 0.385 0.331
LC x ANB Cor 0.364 0.512 0.728 0.686
ANB SD 0.178 0.079 0.026 0.057
s(Age) × ANB SD 0.469 0.144 0.000 0.000
s(Age) × LC SD 0.000 0.868 1.554 0.290
Observation SD 0.403 0.381 0.181 0.166
AIC 995.2 1238.5 -27.0 -180.1
BIC 1062.7 1309.5 40.5 -109.1
Log likelihood -482.587 -604.237 28.499 105.070
REML criteria 965.175 1208.474 -56.997 -210.140
Note. FA = Fractional Anisotropy, AD = Axial Diffusivity, SD = Standard Deviation, Cor = Correlation,
FPT = frontopontine, ANB = Ascending Noradrenergic Bundle, LC = Locus Coeruleus, s.e. = standard
error, t = test statistic.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
81
Table 49: GAM Smooth Portion of Longitudinal GAMM Results

Note. AD = Axial Diffusivity, ANB = Ascending Noradrenergic Bundle, FA = Fractional Anisotropy, LC = Locus
Coeruleus, s.e. = standard error, t = test statistic, q = False discovery rate correction for multiple testing.

Young FA Older FA Young AD Older AD
s(Age) × LC

edf = 1.000 edf = 2.842 edf = 7.027 edf = 2.457

F = 2.336 F = 2.903 F = 4.051 F = 3.186

p = 0.127 p = 0.042 p < 0.001 p = 0.045
q = 0.2 q = 0.069 q < 0.001 q = 0.056
s(Age) × ANB

edf = 3.064 edf = 1.706 edf = 1.000 edf = 1.000

F = 5.473 F = 1.223 F = 3.609 F = 52.638

p < 0.001 p = 0.185 p = 0.058 p < 0.001
q < 0.001 q = 0.2 q = 0.072 q < 0.001
N 666 843 666 843
RMSE 0.58 0.64 0.30 0.28
estimate 0.606 0.605 0.899 0.927
RUNNING HEAD: AGING AND LC DIFFUSIVITY
82
Figure 19: GAMM Longitudinal Fixed Effects Between Age Groups and 95% Confidence
Intervals

Note. ANB = Ascending Noradrenergic Bundle, LC = Locus Coeruleus. Shaded region denotes
95% confidence interval.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
83
Table 50: Linear Mixed Effects Portion of Longitudinal GAMM RAVLT Results

Young FA Older FA Young AD Older AD
Fixed Effects
(Intercept) 0.027 -0.040 0.027 -0.040

s.e. = 0.023 s.e. = 0.026 s.e. = 0.023 s.e. = 0.026

t = 1.158 t = -1.532 t = 1.158 t = -1.532
FPT Linear 1.377 1.563 1.377 1.563

s.e. = 0.027 s.e. = 0.025 s.e. = 0.027 s.e. = 0.025

t = 51.087 t = 63.547 t = 51.087 t = 63.547
FPT Quadratic 0.795 0.710 0.795 0.710

s.e. = 0.024 s.e. = 0.021 s.e. = 0.024 s.e. = 0.021

t = 33.173 t = 33.766 t = 33.173 t = 33.766
Gender: Female 0.008 0.023 0.008 0.023

s.e. = 0.027 s.e. = 0.027 s.e. = 0.027 s.e. = 0.027

t = 0.283 t = 0.839 t = 0.283 t = 0.839
s(RAVLT) × LC -0.012 -0.004 -0.012 -0.004

s.e. = 0.013 s.e. = 0.013 s.e. = 0.013 s.e. = 0.013

t = -0.936 t = -0.299 t = -0.936 t = -0.299
s(RAVLT) × ANB -0.016 -0.022 -0.016 -0.022

s.e. = 0.020 s.e. = 0.020 s.e. = 0.020 s.e. = 0.020

t = -0.811 t = -1.108 t = -0.811 t = -1.108
Random Effects
FPT SD 0.182 0.196 0.182 0.196
FPT x LC Cor 1.000 0.142 1.000 0.142
FPT x ANB Cor 0.569 0.479 0.569 0.479
LC SD 0.033 0.098 0.033 0.098
LC x ANB Cor 0.569 0.465 0.569 0.465
ANB SD 0.349 0.341 0.349 0.341
s(RAVLT) × ANB SD 0.000 0.000 0.000 0.000
s(RAVLT) × LC SD 0.000 0.000 0.000 0.000
Observation SD 0.189 0.165 0.189 0.165
AIC -0.9 -127.1 -0.9 -127.1
BIC 66.2 -56.9 66.2 -56.9
Log likelihood 15.465 78.529 15.465 78.529
REML criteria -30.931 -157.059 -30.931 -157.059
Note. FA = Fractional Anisotropy, AD = Axial Diffusivity, SD = Standard Deviation, Cor = Correlation, FPT =
frontopontine, ANB = Ascending Noradrenergic Bundle, LC = Locus Coeruleus, s.e. = standard error, t = test
statistic, RAVLT = Rey-Auditory Verbal Learning Delay Score.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
84
Table 51: GAM Smooth Portion of Longitudinal GAMM RAVLT Results

Young FA
GAM
Older FA
GAM
Young AD
GAM
Older AD
GAM
s(RAVLT) ×
ANB
estimate estimate estimate estimate

edf = 1.000 edf = 1.000 edf = 1.000 edf = 1.000

F = 0.876 F = 0.089 F = 0.876 F = 0.089

p = 0.350 p = 0.765 p = 0.350 p = 0.765
s(RAVLT) × LC estimate estimate estimate estimate

edf = 1.000 edf = 1.000 edf = 1.000 edf = 1.000

F = 0.659 F = 1.227 F = 0.659 F = 1.227

p = 0.417 p = 0.268 p = 0.417 p = 0.268
N 648 795 648 795
RMSE 0.29 0.28 0.29 0.28
estimate 0.907 0.924 0.907 0.924
Note. AD = Axial Diffusivity, ANB = Ascending Noradrenergic Bundle, FA = Fractional Anisotropy, LC = Locus
Coeruleus, s.e. = standard error, t = test statistic, RAVLT = Rey-Auditory Verbal Learning Delay Score.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
85
Figure 20: GAMM RAVLT Longitudinal Fixed Effects Between Age Groups and 95%
Confidence Intervals

Note. ANB = Ascending Noradrenergic Bundle, LC = Locus Coeruleus, RAVLT = Rey-Auditory Verbal
Learning Delay Score. Shaded region denotes 95% confidence interval.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
86
Chapter 3: LC diffusivity may be influenced by sleep deprivation

1. Introduction

Fractional anisotropy is a non-invasive diffusion MRI metric that reflects the structural
integrity of axons in the brain (Beck et al., 2021; Soares et al., 2013). In most cases, fractional
anisotropy peaks around midlife and declines into old age (Beck et al., 2021; Kiely et al., 2022).
Observations of higher fractional anisotropy in the locus coeruleus (LC) of older adults,
compared with younger adults, has prompted interest in exploring the potential underlying
mechanisms that may explain such findings, (see chapter 2, Porat et al., 2022). Fractional
anisotropy has been reported to reduce in young adults following acute injury or
neurodegeneration, but limited studies have observed such reductions in healthy adults (Beaudet
et al., 2020; Budde et al., 2011; Grambaite et al., 2010; Harris et al., 2016; Kantarci, 2014;
Sexton et al., 2010; Singh et al., 2016). This raises the question of what underlying mechanisms
could result in decreased fractional anisotropy in younger adults relative to older adults.
Recent studies investigating diffusivity and sleep deprivation have observed changes in
fractional anisotropy within a 24-hour period (Elvsåshagen et al., 2015; Voldsbekk et al., 2021;
Voldsbekk et al., 2020). In two studies, fractional anisotropy in young adults increased
throughout the waking day, but decreased after a night of sleep deprivation (Elvsåshagen et al.,
2015; Voldsbekk et al., 2020). A recent study has also reported a positive correlation between
fractional anisotropy and increased sleep quality (Khalsa et al., 2017). Given that the role of the
LC in sleep-to-wake transitions and the LC’s complete inactivity during REM sleep is well
documented (Berridge & Waterhouse, 2003; Gutchess et al., 2020; Hayat et al., 2020; Khanday
RUNNING HEAD: AGING AND LC DIFFUSIVITY
87
et al., 2016; Stock et al., 1981; Takahashi et al., 2010), this raises the question: could a night of
sleep deprivation affect LC diffusivity in young adults differently from older adults?
Generally, older adults accrue sleep problems throughout life that eventually result in
occasional to frequent bouts of sleep loss, so the effects of acute sleep deprivation may not,
initially, be as pronounced when compared to younger adults (Lavoie et al., 2018; Mander et al.,
2016). For instance, adolescents experience greater changes in mood and anxiety after sleep
deprivation than do older adults (Talbot et al., 2010). Likewise, young adults show more
attentional deficits and behavioral impairments after sleep deprivation than do older adults
(Duffy et al., 2009; Zitting et al., 2018).
This is not surprising given that a night of sleep deprivation can increase LC activity which
results in decreased top-down modulation from the frontal cortex and simultaneously increases
LC-amygdala functional connectivity (Goldstein & Walker, 2014; Gong et al., 2021; Mallick &
Singh, 2011). Specifically, we are interested in whether the diffusivity within the LC might
reflect a biomarker for transient states of inflammation or neuronal activity. Animal research
suggests that inflammation restricts fluid flow (Yi et al., 2019).
In addition, neuronal activity tends to swell neurons, reducing the interstitial fluid-filled
spaces between them and decreasing diffusivity (Abe et al., 2017; Iwasa et al., 1980; Le Bihan et
al., 2006; Nunes et al., 2021; Svoboda & Syková, 1991). Sleep deprivation, as well as LC
activity, affects both neuronal activity and inflammation (Grandner et al., 2016; Holth Jerrah et
al., 2019; Irwin et al., 2016; Ju et al., 2017). Here, we utilize the Stockholm SLEEPY Brain data
set, final N = 49, which to our knowledge, is the only publicly available diffusion imaging data
set with sleep deprived young and older adults (Akerstedt, 2016). Of note, the sleep deprivation
paradigm in SLEEPY data set is captured between, not within subjects for the diffusion weighted
RUNNING HEAD: AGING AND LC DIFFUSIVITY
88
scanning portion. Therefore, given the limitations of the data set, we offer initial analyses that we
encourage future studies to replicate with more stringent study designs. Despite the limitations,
we believe our findings are relevant for better understanding age-related differences in LC
fractional anisotropy.
2. Methods

The Stockholm SLEEPY Brain Study aims to investigate sleep deprivation in association
to brain imaging and emotion regulation (Akerstedt, 2016). Participants were instructed to
maintain normal sleep patterns (sleep rested) or go to sleep 3 hours before waking up at their
normal wake times (sleep deprived, e.g., if a participant typically wakes up at 8am, for the sleep
deprivation condition they would go to sleep at 5am but still wake up at 8am), to mimic real-
world sleep deprivation.
Although the SLEEPY Brain study had two time points per participant (one sleep rested,
one sleep deprived), only one time point included a diffusion-weighted scans, and therefore each
participant has diffusion imaging in either the sleep-rested or sleep-deprived condition.
Demographics with final N’s, are displayed in Table 52. Our exclusion criteria included poor
quality diffusion-weighted scans, poor atlas registration of ROIs, or missing data in statistical
analyses. Poor scan and registration quality was determined through visual inspection by two
independent researchers. Table 53 displays the diffusion imaging parameters for the data set.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
89
Table 52: Demographics
Younger Adults Older Adults p
a

Stockholm SLEEPY Brain (SLEEPY)
Age in Years 20-30 65-75
Sex 0.8
Male 6 (38) 15 (45)
Female 10 (62) 18 (55)
Total 16 33
Sleep Deprived 9 (56) 16 (48) 0.8
Note. Age is reported as ranges.
a
Statistical tests performed: chi-square test of independence (for comparisons across
sexes and sleep deprivation conditions); Wilcoxon rank-sum test (for age).
b
Statistics presented: Mean (SD).
c
Statistics presented: n (% of total)

Table 53: Diffusion MRI Sequence Parameters
Item Duration or Size
Stockholm SLEEPY Brain
Scanner 3-Tesla General Electric Discovery
Head coil 8-channel
Sequence Sagittal plane and five volumes collected
without diffusion weighting (b = 0).
5:55 minutes
Parameters Repetition Time 7000ms
Echo Time 80ms
Gradient Directions 45
Diffusion Weighting b = 1,000 s/mm
2

Bandwidth 1502 Hz/pixel
FOV 220mm
Slice Thickness 2.3mm isotropic
Other Diffusion-weighted imaging was only collected during the second
scanning session of this study, yielding a total of 26 sleep deprived
and 27 non sleep deprived subjects, randomly assigned. For more
information please see (Nilsonne et al., 2016)
RUNNING HEAD: AGING AND LC DIFFUSIVITY
90
2.1.DWI Processing

Preprocessing of diffusion imaging scans was conducted with custom University of
Southern California’s Laboratory of Neuroimaging (LONI) pipelines. FSL (v6.3) eddy-current
and motion correction as well as brain extraction tool, resampling to isotropic resolution of
2mm
3
were applied for each subject

(Dinov et al., 2009; Smith et al., 2004). To calculate
fractional anisotropy (FA), mean diffusivity (MD) and radial diffusivity (RD), we used MRtrix
(v3.1) (Tournier et al., 2019). Dahl’s LC meta-masks were used to measure FA and diffusivity
within the left and right LC of each participant (Dahl et al., 2021). The right and left ascending
noradrenergic bundle atlases were adopted from Tang et al. (2020). Lastly, the right and left
frontopontine tract atlases were used as a control region, given its location ventral to the LC
(Tang et al., 2018). We display all three ROIs on a single brain, MNI152 linear 1mm resolution
space, in Figure 21.
To analyze fractional anisotropy and diffusion within subject native space we first
registered FA images and each atlas into MNI152 linear, 1mm brain space. Then, the atlases
were backwarped into native space with K-nearest neighbor interpolation using ANTS nonlinear
registration (Avants et al., 2008; Sun et al., 2020). We developed a custom MATLAB script to
visualize registration of atlases and FA images for each subject (MATLAB ver. R2019a). Two
individual researchers independently confirmed accurate subject atlas registration (any
disagreement resulted in subject exclusion from analyses). We then thresholded and converted
the subject-registered atlases into a binary masks. Lastly, we multiplied the FA, mean, or radial
diffusion images by each ROI binary mask to obtain respective diffusion values within each
mask, per voxel, within native space. FA or diffusivity values were then averaged to provide one
value per participant per ROI, in each hemisphere.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
91
2.2.Statistical Analyses

All statistical analyses were conducted using the R software (Team, 2020) with several
statistical packages (Ahlmann-Eltze, 2019; Aust & Barth, 2020; Kassambara; Lenth, 2021;
Sjoberg et al., 2021; Wicham, 2017; Wickham, 2016; Xie, 2021). Fractional anisotropy, mean
diffusivity, and radial diffusivity values in the LC, ascending noradrenergic bundle, and
frontopontine tract were analyzed within each dataset using a 2 × 2 × 3 × 2 factorial design in
which age (younger, older) and sleep condition (rested, deprived) were between-subject factors
and ROI (noradrenergic bundle, locus coeruleus, frontopontine tract) and hemisphere (left, right)
were repeated-measures factors. Post-hoc analyses included Wilcoxon t-tests between sleep
rested and sleep deprived groups.
Effect sizes are reported with 𝜂̂ !
"
(generalized eta squared) for ANOVA effects and we
provide 95% confidence intervals to compare between means. Levene’s tests were used in R to
explore ANOVA assumptions of equal variances. Greenhouse-Geisser correction was
automatically applied for ANOVA departures from sphericity. This paper focuses on fractional
anisotropy to follow up on our prior age difference findings (Chapter 2; Porat et al., 2022), but
we include mean and radial diffusivity findings in the supplementary material.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
92
Figure 21: The ROI atlases of the Locus Coeruleus, Noradrenergic Bundle, and Frontopontine
Tract

Note. Figure 1 displays the locus coeruleus (yellow), noradrenergic bundle (red), and frontopontine (blue)
tracts registered to MNI152 space. The noradrenergic bundle is one continuous bundle (part of the
temporal lobe segment is not pictured).

3. Results

3.1.Fractional anisotropy in the LC is higher in older adults, relative to younger adults

In the SLEEPY Brain dataset, the Age x ROI interaction did not quite reach significance,
F(1.76, 72.17) = 3.09, p = .058, 𝜂̂ !
"
= 0.03, Table 3. Likewise, the Age x ROI x Sleep Condition
interaction did not quite achieve significance, F(1.76, 72.17) = 3.12, p = .056, 𝜂̂ !
"
= 0.031. Here,
we report planned contrasts of fractional anisotropy across ROIs between age groups and sleep
conditions. Means and confidence intervals for the SLEEPY dataset are available in Table 54,
RUNNING HEAD: AGING AND LC DIFFUSIVITY
93
and plots for between and within age group differences are shown in Figures 22 and 23,
respectively.
In the sleep rested condition, we observed no significant differences in fractional
anisotropy between young and older adults in the left or right LC, t(41) = 0.181, p = .857, and
t(41) = 1.469, p = .149, respectively. We observed no significant differences in fractional
anisotropy between sleep rested young and older adults in the left noradrenergic bundle, t(41) =
.619, p = .539, but significantly higher fractional anisotropy in the right noradrenergic bundle of
sleep rested younger adults, compared to older, t(41) = 2.56, p = .014. Similarly, we observed
no significant differences in fractional anisotropy between sleep rested young and older adults in
the left frontopontine tract, t(41) = 0.838, p = .407, but significantly higher fractional anisotropy
in the right frontopontine tract of sleep rested younger adults, compared to older, t(41) = 2.161, p
= .036.
Within the sleep deprived condition, younger adults had lower fractional anisotropy in
the LC, compared with older adults, that was not significant in the left locus coeruleus, t(41) = -
1.658, p = .104, but was significant in the right locus coeruleus, t(41) = -2.451, p = .018. We also
observed significantly higher fractional anisotropy in sleep deprived younger adults, compared
with older adults, in both the left and right noradrenergic bundle, t(41) = 3.286, p = .002 and
t(41) = 4.291, p < .001, respectively. No significant differences were observed between younger
and older adult’s fractional anisotropy in the sleep deprived conditions in the left or right
frontopontine tract, t(41) = 0.163, p = .870, and t(41) = 0.296, p = .769, respectively.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
94
Table 54: SLEEPY Fractional Anisotropy Mixed ANOVA
Variable 𝐹 𝑑𝑓
!
""
𝑑𝑓
#
""
𝑀𝑆𝐸 𝑝 𝜂̂ "
#

Age (Younger, Older) 1.32 1 41 0.01 .257 .011
Sleep Condition (Rested, Deprived) 1.19 1 41 0.01 .282 .010
ROI (LC, Noradrenergic Bundle, Frontopontine) 115.99 1.76 72.17 0.00 <.001 .540
Hemisphere (Left, Right) 15.92 1 41 0.00 <.001 .040
Age × Sleep Condition 2.44 1 41 0.01 .126 .020
Age × ROI 3.09 1.76 72.17 0.00 .058 .030
Sleep Condition × ROI 2.48 1.76 72.17 0.00 .098 .024
Age × Hemisphere 1.27 1 41 0.00 .267 .003
Sleep Condition × Hemisphere 1.22 1 41 0.00 .275 .003
ROI × Hemisphere 18.37 1.57 64.45 0.00 <.001 .054
Age × Sleep Condition × ROI 3.12 1.76 72.17 0.00 .056 .031
Age × Sleep Condition × Hemisphere 2.08 1 41 0.00 .157 .005
Age × ROI × Hemisphere 0.34 1.57 64.45 0.00 .659 .001
Sleep Condition × ROI × Hemisphere 0.24 1.57 64.45 0.00 .732 .001
Age × Sleep Condition × ROI × Hemisphere 0.24 1.57 64.45 0.00 .731 .001

RUNNING HEAD: AGING AND LC DIFFUSIVITY
95
Table 55: SLEEPY Fractional Anisotropy Means, Standard Error, Degrees of Freedom and 95%
Confidence Intervals
SLEEPY

Locus Coeruleus Noradrenergic Bundle Frontopontine Tract
Young
Adult
Older
Adult
Young
Adult
Older
Adult
Young
Adult
Older
Adult
Left Hemisphere, Rested
M
a
0.405 0.399 0.323 0.315 0.475 0.446
SE 0.027 0.019 0.01 0.007 0.029 0.02
Lower CI 0.35 0.36 0.303 0.301 0.417 0.405
Upper CI 0.46 0.437 0.344 0.33 0.533 0.486
Left Hemisphere, Deprived
M
a
0.343 0.395 0.35 0.311 0.469 0.463
SE 0.025 0.018 0.009 0.007 0.027 0.019
Lower CI 0.292 0.358 0.331 0.298 0.415 0.424
Upper CI 0.394 0.432 0.369 0.325 0.523 0.503
Right Hemisphere, Rested
M
a
0.435 0.394 0.272 0.249 0.49 0.415
SE 0.023 0.016 0.007 0.005 0.028 0.02
Lower CI 0.389 0.361 0.257 0.239 0.433 0.375
Upper CI 0.482 0.427 0.287 0.26 0.547 0.455
Right Hemisphere, Deprived
M
a
0.339 0.404 0.282 0.246 0.44 0.43
SE 0.021 0.016 0.007 0.005 0.026 0.019
Lower CI 0.295 0.372 0.268 0.236 0.387 0.392
Upper CI 0.382 0.436 0.296 0.256 0.493 0.469
Note. CI = confidence interval.
a
degrees of freedom = 41.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
96
Figure 22: SLEEPY Brain Fractional Anisotropy in Left and Right Locus Coeruleus and
Noradrenergic Bundles in Younger and Older Adults

Note. Figure 2 displays fractional anisotropy between younger and older adults in either the sleep
rested or sleep deprived conditions. In the left (A) and right (B) locus coeruleus, we observed no
significant differences in fractional anisotropy between young and older adults in the sleep rested
condition. However, in the sleep deprived condition, younger adults displayed significantly
lower fractional anisotropy, compared to older adults. In the left noradrenergic bundle (C) we
observed no significant differences between young and older adult fractional anisotropy in the
sleep rested condition. In the sleep deprived condition (C – row 2), younger adults had
significantly higher fractional anisotropy, relative to older adults. In the right noradrenergic
bundle (D) younger adults, compared to older adults, had significantly higher fractional
anisotropy in the sleep rested condition. Fractional anisotropy differences were more pronounced
in the sleep deprived condition.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
97
Figure 23: SLEEPY Brain Fractional Anisotropy in Left and Right Locus Coeruleus and
Noradrenergic Bundles between Younger and Older Adults

Note. Figure 3 displays fractional anisotropy between sleep rested or sleep deprived young and
older adults. In the left locus coeruleus (A) and right locus coeruleus (B) we observed
significantly higher fractional anisotropy in sleep rested young adults, compared to sleep
deprived young adults. We observed no significant difference of fractional anisotropy between
sleep rested and sleep deprived older adults in the left or right locus coeruleus. In the left
noradrenergic bundle (C) and right noradrenergic bundle (D) we observed no significant
difference in fractional anisotropy in sleep rested or sleep deprived younger adults; nor sleep
rested or sleep deprived older adults.

4. Discussion

Here, using the publicly available SLEEPY dataset (Akerstedt, 2016), we examined the
age-related diffusivity of the LC, ascending noradrenergic bundle, and frontopontine tracts. We
did not observe FA differences in the locus coeruleus between young and older adults, who were
RUNNING HEAD: AGING AND LC DIFFUSIVITY
98
sleep rested, but only those who were sleep deprived. We did, however, observe typical age
differences in diffusivity of the noradrenergic bundle between sleep rested and sleep deprived
young and older adults. Post-hoc analyses of the SLEEPY dataset revealed significant
differences in LC fractional anisotropy between sleep rested and sleep deprived young adults,
but not between sleep rested and sleep deprived older adults.
While fractional anisotropy was higher in older adults relative to younger adults within
the LC itself (Porat et al., 2022), older adults typically showed lower fractional anisotropy than
younger adults in the noradrenergic bundle ascending tract, a typical age-related pattern observed
in white matter (Medina & Gaviria, 2008; Sibilia et al., 2017; Sullivan & Pfefferbaum, 2006;
Voineskos et al., 2012). Mean and radial diffusivity in the LC also showed some age differences
(results and tables provided in the supplementary material), although not as pronounced as
fractional anisotropy.
We observed expected age differences in the LC of sleep rested adults, with lower mean
and radial diffusivities in younger adults, relative to older adults. In the sleep deprived condition,
we did not observe significant differences in LC mean and radial diffusivities between age
groups. However, we observed higher LC mean and radial diffusivity in younger adults when
sleep deprived, compared to sleep rested young adults (see supplementary tables). Meanwhile,
levels of diffusivity within older adults between sleep rested and sleep deprived conditions
remained relatively unchanged. This raises the question of what underlying neural mechanisms
may contribute to lower FA in sleep deprived younger adults but not older adults.
We offer some speculation based on studies involving the LC and glymphatic system, the
brain’s amyloid clearance system during sleep (Benveniste et al., 2019). Studies have identified
the locus coeruleus as highly active during wake states and virtually silent during REM sleep
RUNNING HEAD: AGING AND LC DIFFUSIVITY
99
(Berridge & Waterhouse, 2003; Gong et al., 2021; Gutchess et al., 2020; Hayat et al., 2020;
Khanday et al., 2016; Takahashi et al., 2010). Specifically, LC-specific acid-sensing ion channels
may modulate sleep-wake state transitions (Mir & Jha, 2021). When rats are administered
dexmedetomidine, an anesthetic that decreases LC firing rate and induces light sleep, the
glymphatic system was significantly more enhanced in comparison with rats receiving
isoflurane, which does not inhibit norepinephrine release (Benveniste et al., 2017).
Norepinephrine has been reported to modulate extracellular space through structural
changes in astrocytes (Sherpa et al., 2017). Astrocytes are an important aspect of the glymphatic
system and sleep regulation (Benveniste et al., 2019; Haydon, 2017; Jessen et al., 2015; Smith et
al., 2017), as AQP4 are astrocytic end feet responsible for the glymphatic system’s movement of
fluid (Benveniste et al., 2019; Jessen et al., 2015). Astrocytic activity has also been associated
with reduced fractional anisotropy in neurodegenerative disorders and injury (Beppu et al., 2003;
Kato et al., 2021). And recently, acute injury in swine was reported to result in lower fractional
anisotropy that correlated with increased astrocytic swelling, specifically those of aquaporin-4
(AQP4) channels (Lee et al., 2021).
Does increased LC activity during a night of sleep deprivation result in decreased
fractional anisotropy in younger adults by recruiting astrocytic activity? While specific studies
are needed to address this hypothesis, some studies have reported positive correlations between
increased LC activation and astrocytic activity (Giorgi et al., 2020; Nakamoto et al., 2017).
Specifically, norepinephrine has been associated with increased astrogliosis (Giorgi et al., 2020;
Paukert et al., 2014) and chronic sleep deprivation in mice was reported to increase astrocytic
coverage of axons and synapses (Bellesi et al., 2015; DiNuzzo & Nedergaard, 2017).
RUNNING HEAD: AGING AND LC DIFFUSIVITY
100
Though limited, some studies have reported a reduction in astrocytic processes in aging
mice (Popov et al., 2021), a causal link between senescent astrocyte accumulation in aging mice
and Alzheimer’s-related tau pathology (Bussian et al., 2018), and that astrocytic impairment in
Alzheimer’s disease is associated with decreased amyloid clearance (Iram et al., 2016). This
raises the question: could astrocytic dysfunctions associated with aging contribute to the lack of
sleep-deprived associations with fractional anisotropy in the LC of older adults? Notably, studies
suggest that dysregulation of astrocytic modulated adenosine receptors in aging mice and
humans may result in poor homeostatic regulation of sleep (Bellesi et al., 2017; Florian et al.,
2011; Halassa et al., 2009; Mander et al., 2017).
Early Alzheimer’s pathology is also associated with poor sleep and reduced fractional
anisotropy years before cognitive deficits appear (Benveniste et al., 2019; Cedernaes et al., 2017;
Ehrenberg et al., 2018; Kantarci, 2014; Kantarci et al., 2017; Li et al., 2016; Mander et al., 2016;
Molinuevo et al., 2014; Rose et al., 2008; Ryu et al., 2017; Stebbins & Murphy, 2009). Some
studies have reported increases in fractional anisotropy post-acute injury (Harris et al., 2016), but
the type of astrocytic molecular activity varies by injury (Budde et al., 2011; Lee et al., 2021).
Several studies have argued that astrocytes and other glial cells affect diffusion MRI
(Blumenfeld-Katzir et al., 2011; Budde et al., 2011; Singh et al., 2016). And a recent study found
decreased fractional anisotropy is associated with gliosis, reporting greater significance in gray
matter regions, relative to white matter (Chary et al., 2021). Together, these studies could
suggest an interplay between LC modulated astrocytic activity and poor sleep may be affecting
diffusion in the brain.
We encourage other labs to replicate our findings in a larger sample size and improved
MRI study design. Mouse models can be used to study fractional anisotropy and the in-vivo
RUNNING HEAD: AGING AND LC DIFFUSIVITY
101
effects of A1 (pro-inflammatory) and A2 (anti-inflammatory) astrocyte reactivity during sleep
deprivation. Future studies may provide clinical insights by sleep depriving a larger sample of
young and older adults. REM specific sleep-deprivation in a diffusion imaging study in
combination with proton magnetic resonance spectroscopy (H-MRS) to assess astrocytic
metabolism may shed light on age-related relationships between LC fractional anisotropy and
sleep deprivation. Studying the relationships between sleep and fractional anisotropy may
ultimately shed light in differences in the underlying mechanisms that result in structural and
behavioral changes between young and older adults.
4.1.Limitations

The SLEEPY dataset is relatively underpowered given the low sample size between sleep
rested and sleep deprived groups and therefore our results should be considered for replication in
a larger sleep-deprived dataset. However, this dataset is the only available sleep-diffusion public
dataset with an older adult sample, so these initial findings may provide preliminary results for
future studies. We did not examine axial diffusivity, which may be influenced by sleep
deprivation and should be further explored in the future (Elvsåshagen et al., 2015).
4.2.Conclusions

To our knowledge, this is the first study to compare diffusivity differences in-vivo in the
locus coeruleus and noradrenergic bundle in sleep deprived young and older adults. We
encourage researchers to attempt to replicate our findings in a larger sample as the associations
between sleep deprivation and diffusivity in the locus coeruleus may be relevant for clinical
applications such as detection of sleep disorders that may have significant long-term brain
effects. Further exploration of sleep and diffusivity in neurodegenerative diseases may allow for
RUNNING HEAD: AGING AND LC DIFFUSIVITY
102
DTI to be used as a predictive biomarkers as sleep disruptions often predate dementia diagnoses
by many years.
5. Supplementary Material

We observed significant Age x Gender x ROI x Hemisphere interactions within LEMON
mean diffusivity F(1.41, 304.60) = 4.74, p = .019, 𝜂̂ !
"
= .004, and radial diffusivity, F(1.46,
314.29) = 5.22, p = .012, 𝜂̂ !
"
= .005. For the SLEEPY Brain dataset, we observed a trending Age
x ROI x Hemisphere x Sleep Condition interaction for mean and radial diffusivities, F(1.25,
51.27) = 3.45, p = .06, 𝜂̂ !
"
= .011, and F(1.28, 52.65) = 2.65, p = .101, 𝜂̂ !
"
= .008, respectively.
Means and confidence intervals for the SLEEPY dataset mean and radial diffusivities are
available in Tables 56 and 57.
In the sleep rested condition, we observed no significant differences in mean or radial
diffusivities between young and older adults in the left locus coeruleus, t(41) = -1.134, p = .263,
and t(41) = -1.012, p = .317, respectively. However, in the right locus coeruleus, younger adults
had significantly lower mean and radial diffusivities, compared with older adults, t(41) = -2.543,
p = .014, and t(41) = -2.345, p = .023, respectively. Both mean and radial diffusivities in the left
noradrenergic bundle t(41) = -2.283, p = .027, and t(41) = -2.122, p = .039, respectively, and
mean and radial diffusivities in the right noradrenergic bundle t(41) = -3.185, p = .002, and t(41)
= -3.417, p = .001, were significantly lower in younger adults, compared to older adults.
Similarly, mean, and radial diffusivities in the left frontopontine tract, t(41) = -2.393, p = .021,
and t(41) = -2.28, p = .027, and right frontopontine tract, t(41) = -2.595, p = .013, and t(41) = -
2.593, p = .013, were significantly lower in younger adults, compared to older adults.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
103
Within the sleep deprived condition, we observed no significant differences in mean or
radial diffusivities between young and older adults in the left locus coeruleus, t(41) = 0.177, p =
.860, and t(41) = 0.666, p = .509, respectively. Nor did we observe significant differences
between young and older adults' mean and radial diffusivities in the right locus coeruleus, t(41)
= 0.949, p = .348, and t(41) = 1.491, p = .143, respectively. Both mean and radial diffusivities in
the left noradrenergic bundle t(41) = -2.371, p = .022, and t(41) = -2.728, p = .009, respectively,
and mean and radial diffusivities in the right noradrenergic bundle t(41) = -2.859, p = .006, and
t(41) = -3.265, p = .002, were significantly lower in younger adults, compared to older adults. No
significant differences in mean or radial diffusivities of the left frontopontine tract, t(41) = -
0.882, p = .387, and t(41) = -0.771, p = .444, and right frontopontine tract, t(41) = -1.679, p =
.101, and t(41) = -1.588 p = .120, were observed between younger and older adults.
Table 56: SLEEPY Mean Diffusivity Mixed ANOVA
Variable 𝐹 𝑑𝑓
!
""
𝑑𝑓
#
""
𝑀𝑆𝐸 𝑝 𝜂̂"
#

Age (Younger, Older) 11.07 1 41 0.00 .002 .074
Sleep Condition (Rested, Deprived) 1.45 1 41 0.00 .235 .010
ROI (LC, Noradrenergic Bundle, Frontopontine) 296.37 1.61 66.05 0.00 <.001 .787
Hemisphere (Left, Right) 10.93 1 41 0.00 .002 .017
Age × Sleep Condition 4.55 1 41 0.00 .039 .032
Age × ROI 0.43 1.61 66.05 0.00 .610 .005
Sleep Condition × ROI 5.02 1.61 66.05 0.00 .014 .059
Age × Hemisphere 2.89 1 41 0.00 .097 .005
Sleep Condition × Hemisphere 1.81 1 41 0.00 .186 .003
ROI × Hemisphere 21.67 1.25 51.27 0.00 <.001 .063
Age × Sleep Condition × ROI 1.98 1.61 66.05 0.00 .154 .024
Age × Sleep Condition × Hemisphere 2.41 1 41 0.00 .129 .004
Age × ROI × Hemisphere 0.22 1.25 51.27 0.00 .699 .001
RUNNING HEAD: AGING AND LC DIFFUSIVITY
104
Sleep Condition × ROI × Hemisphere 0.75 1.25 51.27 0.00 .417 .002
Age × Sleep Condition × ROI × Hemisphere 3.45 1.25 51.27 0.00 .060 .011

Table 57: SLEEPY Radial Diffusivity Mixed ANOVA
Variable 𝐹 𝑑𝑓
!
""
𝑑𝑓
#
""
𝑀𝑆𝐸 𝑝 𝜂̂"
#

Age (Younger, Older) 8.78 1 41 0.00 .005 .061
Sleep Condition (Rested, Deprived) 1.48 1 41 0.00 .231 .011
ROI (LC, Noradrenergic Bundle, Frontopontine) 179.49 1.69 69.15 0.00 <.001 .685
Hemisphere (Left, Right) 13.46 1 41 0.00 .001 .024
Age × Sleep Condition 5.02 1 41 0.00 .031 .036
Age × ROI 1.11 1.69 69.15 0.00 .328 .013
Sleep Condition × ROI 5.54 1.69 69.15 0.00 .009 .063
Age × Hemisphere 2.22 1 41 0.00 .144 .004
Sleep Condition × Hemisphere 2.09 1 41 0.00 .156 .004
ROI × Hemisphere 23.83 1.28 52.65 0.00 <.001 .069
Age × Sleep Condition × ROI 2.58 1.69 69.15 0.00 .092 .030
Age × Sleep Condition × Hemisphere 1.70 1 41 0.00 .200 .003
Age × ROI × Hemisphere 0.22 1.28 52.65 0.00 .704 .001
Sleep Condition × ROI × Hemisphere 0.62 1.28 52.65 0.00 .473 .002
Age × Sleep Condition × ROI × Hemisphere 2.65 1.28 52.65 0.00 .101 .008

RUNNING HEAD: AGING AND LC DIFFUSIVITY
105
Table 58: SLEEPY Mean Diffusivity Means, Standard Error, Degrees of Freedom and 95%
Confidence Intervals
SLEEPY Locus Coeruleus Noradrenergic Bundle Frontopontine Tract
Young Adult Older Adult Young Adult Older Adult Young Adult Older Adult
Left Hemisphere Rested
M
a
0.00165 0.00178 0.00083 9.00E-04 0.00083 0.00105
SE 1.00E-04 7.00E-05 2.00E-05 2.00E-05 7.00E-05 5.00E-05
Lower CI 0.00145 0.00164 0.00078 0.00086 0.00068 0.00094
Upper CI 0.00184 0.00192 0.00088 0.00093 0.00098 0.00115
Left Hemisphere Deprived
M
a
0.00186 0.00184 0.00083 0.00089 0.00083 0.00091
SE 9.00E-05 7.00E-05 2.00E-05 2.00E-05 7.00E-05 5.00E-05
Lower CI 0.00167 0.0017 0.00078 0.00086 0.00069 0.00081
Upper CI 0.00204 0.00197 0.00087 0.00092 0.00097 0.00101
Right Hemisphere Rested
M
a
0.00146 0.00183 0.00098 0.00114 8.00E-04 0.00101
SE 0.00012 8.00E-05 4.00E-05 3.00E-05 7.00E-05 5.00E-05
Lower CI 0.00122 0.00166 9.00E-04 0.00108 0.00066 0.00092
Upper CI 0.0017 0.002 0.00106 0.0012 0.00093 0.00111
Right Hemisphere Deprived
M
a
0.00195 0.00182 0.00101 0.00114 0.00078 0.00091
SE 0.00011 8.00E-05 4.00E-05 3.00E-05 6.00E-05 5.00E-05
Lower CI 0.00172 0.00166 0.00093 0.00109 0.00066 0.00082
Upper CI 0.00217 0.00198 0.00108 0.0012 0.00091 0.00101
Note. CI = confidence interval.
a
degrees of freedom = 41.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
106
Table 59: SLEEPY Radial Diffusivity Means, Standard Error, Degrees of Freedom and 95%
Confidence Intervals
SLEEPY Locus Coeruleus Noradrenergic Bundle Frontopontine Tract
Young Adult Older Adult Young Adult Older Adult Young Adult Older Adult
Left Hemisphere Rested
M
a
0.00126 0.00139 0.00068 0.00075 0.00063 0.00085
SE 1.00E-04 7.00E-05 2.00E-05 2.00E-05 8.00E-05 5.00E-05
Lower CI 0.00106 0.00124 0.00063 0.00071 0.00048 0.00074
Upper CI 0.00147 0.00153 0.00073 0.00078 0.00079 0.00095
Left Hemisphere Deprived
M
a
0.0015 0.00142 0.00067 0.00074 0.00063 7.00E-04
SE 9.00E-05 7.00E-05 2.00E-05 2.00E-05 7.00E-05 5.00E-05
Lower CI 0.00131 0.00129 0.00062 0.00071 0.00049 0.00059
Upper CI 0.00169 0.00156 0.00071 0.00077 0.00077 8.00E-04
Right Hemisphere Rested
M
a
0.0011 0.00141 0.00084 0.001 6.00E-04 0.00082
SE 0.00011 8.00E-05 4.00E-05 3.00E-05 7.00E-05 5.00E-05
Lower CI 0.00088 0.00126 0.00077 0.00095 0.00046 0.00072
Upper CI 0.00132 0.00157 0.00092 0.00106 0.00074 0.00092
Right Hemisphere Deprived
M
a
0.00159 0.0014 0.00087 0.00101 6.00E-04 0.00073
SE 1.00E-04 7.00E-05 4.00E-05 3.00E-05 7.00E-05 5.00E-05
Lower CI 0.00139 0.00126 0.00079 0.00096 0.00047 0.00063
Upper CI 0.00179 0.00155 0.00094 0.00106 0.00073 0.00082
Note. CI = confidence interval.
a
degrees of freedom = 41.
Chapter 4: General Discussion

LC structural integrity is essential to healthy aging (Mather, 2021; Mather & Harley,
2016). However, scientists have only recently examined the locus coeruleus in humans, in-vivo.
In this dissertation, using diffusion MRI, we examined the microstructural properties of the LC
RUNNING HEAD: AGING AND LC DIFFUSIVITY
107
by measuring fractional anisotropy and diffusivity (Beck et al., 2021; Kiely et al., 2022; Lazari &
Lipp, 2021). We used four publicly available data sets to address the following aims: 1) whether
we can replicate observations of higher fractional anisotropy in the LC of older adults, relative to
younger adults; 2) whether LC diffusivity follows a linear or nonlinear trend from early
childhood to late life; and 3) whether there is a similar or different association between sleep
deprivation and LC diffusion between young and older adults.
In brief, we observed non-linear and atypical age-related changes in LC fractional
anisotropy and diffusivity relative to its ascending noradrenergic bundle and the frontopontine
tract. Our results also align with previous observations that higher LC fractional anisotropy is
observed in older adults, relative to young adults. Lastly, we observed lower fractional
anisotropy in the LC of sleep deprived young adults, but not sleep deprived older adults. We
speculate that sleep deprivation may affect LC fractional anisotropy in younger adults differently
than older adults. Together, this dissertation provides evidence to suggest that age-related LC
microstructural properties may be unique, compared to other brain regions.
1. Main Finding 1: Fractional anisotropy in the LC of younger adults is lower
when compared with older adults

In Chapter 1, we aimed to replicate previous reporting of higher fractional anisotropy in
the LC of older adults, relative to younger adults, in a larger sample size (Langley et al., 2020).
With aging, fractional anisotropy in white matter is observed to reduce over time and be lower in
older adults, compared with younger adults (Beaudet et al., 2020; Beck et al., 2021; Bender et
al., 2019; Clewett et al., 2014; Grambaite et al., 2010; Li et al., 2016; Ryu et al., 2017; Zuurbier
et al., 2013). Neurodegenerative diseases exacerbate age-related increases in diffusivity (mean,
RUNNING HEAD: AGING AND LC DIFFUSIVITY
108
radial, axial) and reductions in fractional anisotropy even more (Kantarci, 2014; Kantarci et al.,
2017; Mielke et al., 2012; Molinuevo et al., 2014; Oouchi et al., 2007; Rose et al., 2008).
Though limited, studies have reported that gray matter regions also follow similar trends
as white matter age-related diffusivity, with decreases in fractional anisotropy and increases in
mean diffusivity associated with aging (Köhncke et al., 2021; Weston et al., 2015). So higher
fractional anisotropy in the LC of older adults, relative to young adults, is a curious finding
(Langley et al., 2020). This is also striking because the LC is the first site of abnormal tau
pathology, a hallmark of Alzheimer’s disease (Braak et al., 2011). Gray matter regions
associated with Alzheimer’s disease, like the hippocampus, display reductions in fractional
anisotropy along with increases in mean and radial diffusivities that predate white matter
changes (Jacobs et al., 2013; Weston et al., 2015).
In Study 1, we attempted to expand LC diffusivity findings by investigating diffusion in
the LC’s ascending noradrenergic bundle (ANB). The ANB is implicated in Alzheimer’s disease
and originates in superior portions of the LC where it projects to the entorhinal cortex (Braak et
al., 2011; Sun et al., 2020). For this, we used two publicly available datasets, the Berlin Aging
Study-II (BASEII) and Leipzig Emotion Mind and Body (LEMON), that together include over
500 young and older adults (Babayan et al., 2019; Delius et al., 2015). We chose the
frontopontine tracts as our control region given its ventral location to the LC and lack of
structural change throughout aging (Tang et al., 2018). We used publicly available masks to
calculate diffusion in the LC, ANB, and frontopontine tract.
Specifically, fractional anisotropy, mean, and radial diffusivity in the LC were measured
by adapting an LC meta-mask by Dahl et al., which was created by aggregating LC atlases across
several studies (Dahl et al., 2022). The ANB and frontopontine masks were adapted from a
RUNNING HEAD: AGING AND LC DIFFUSIVITY
109
publicly available archives and derived from the high-resolution human connectome dataset (Sun
et al., 2020; Tang et al., 2018). Additionally, due to the overlap between the LC and ANB, we
conducted along-tract analyses to compare diffusivity along 50 equidistant points that together
create the ANB. Multi-variate ANOVAs and post-hoc t-tests were conducted to examine
differences in diffusivity between young and older adults (Porat et al., 2022).
In Chapter 1, we report significant differences between the LC, ANB, and frontopontine
fractional anisotropy, mean, and radial diffusivities. Specifically, we replicated Langley et al.’s,
findings of higher fractional anisotropy in the LC of older adults, relative to younger adults.
However, older adults displayed reduced fractional anisotropy in the ANB, compared with
younger adults (Porat et al., 2022). We observed no significant differences in the fractional
anisotropy of the frontopontine tract between age groups. For the along-tract analyses of the
ANB, we observed significant reductions in fractional anisotropy in younger adults, relative to
older adults, in the first 10 equidistant points, which approximates the region of the LC (Porat et
al., 2022; Sun et al., 2020). And around points 30-50, which approximates the region of the
entorhinal cortex, we observed higher fractional anisotropy in younger adults, relative to older
adults (Porat et al., 2022).
In summary, we successfully replicated the findings of Langley et al., observing
decreased fractional anisotropy in the LC of older adults, compared with young adults, in two
separate large data sets (Langley et al., 2020; Porat et al., 2022). However, these findings lead to
more questions. Specifically, given that the LC is a gray matter nucleus, are there differences in
fractional anisotropy or diffusivity between gray and white matter regions that may explain our
findings? For example, gray matter diffusion in the hippocampus, in the same BASE-II cohort
RUNNING HEAD: AGING AND LC DIFFUSIVITY
110
was reported with higher mean diffusivity to be positively correlated with aging (Köhncke et al.,
2021).
Recent studies calculating standard brain diffusion tensors (like fractional anisotropy and
mean diffusivity) as well as more advanced techniques (like Neurite Orientation Dispersion and
Density Imaging (NODDI)), have argued that gray matter diffusivity can provide insight into
understanding microstructural properties in regions of interests (Assaf, 2019; Radhakrishnan et
al., 2022). Expected age-related trends in gray matter diffusivity are also observed in cognitively
impaired adults (Kantarci et al., 2005; Müller et al., 2005). Increased mean diffusivity is
observed in the hippocampus of mild cognitive impaired adults relative to healthy controls
(Müller et al., 2005). Studies suggest that gray matter diffusivity provides better sensitivity to
neurodegenerative-related changes in diffusion than white matter diffusivity (Douaud et al.,
2013). Lastly, fractional anisotropy may be associated with cellular composition of gray matter
and agranular gray matter regions may display lower fractional anisotropy, relative to granular
layers of gray matter, though more information is needed for subcortical and brainstem regions
(Assaf, 2019).
One study that investigated deep gray matter diffusivity reported increased fractional
anisotropy in subcortical and brainstem gray matter regions of older adults, relative to younger
adults (Pfefferbaum & Sullivan, 2003). Pfefferbaum et al., speculated that increased iron deposits
are associated with higher fractional anisotropy in older adults relative to younger adults
(Pfefferbaum & Sullivan, 2003). These findings were replicated when Langley et al., observed
reduced diffusion and increased fractional anisotropy in the substantia nigra of older adults,
relative to younger adults (Langley et al., 2020). But Langley et al., argue that iron deposition
cannot influence LC diffusion due to neuromelanin’s complex chemical bonds with copper and
RUNNING HEAD: AGING AND LC DIFFUSIVITY
111
other metals (Langley et al., 2020). Given the uncertainty in the literature, more research is
needed to understand the relationship of LC diffusivity and another question is raised: does
higher LC diffusivity in older adults provide any cognitive benefits?
In the original Langley et al., study, they report increased LC mean, radial, and axial
diffusivity are associated with better RAVLT delayed recall score performance, though no
fractional anisotropy associations were reported (Langley et al., 2020). There were no
associations between LC diffusivity and RAVLT delay scores observed in young adults either
(Langley et al., 2020). This is unexpected given the relationship between diffusion and memory
reported in the literature, even in gray matter regions, is that higher diffusion is associated with
worse memory outcomes (Bender et al., 2019; Blumenfeld-Katzir et al., 2011; den Heijer, der
Lijn, Vernooij, de Groot, Koudstaal, van der Lugt, et al., 2012; Douaud et al., 2013; Han et al.,
2016; Kantarci, 2014; Sibilia et al., 2017; Venkatesh et al., 2020; Voineskos et al., 2012). For
example, studies report decreased fractional anisotropy in the hippocampus of older adults,
relative to younger adults – and no significant associations between memory performance and
fractional anisotropy in older adults (Venkatesh et al., 2020). Age-expected increases in
hippocampal mean diffusivity have been associated with poor episodic memory in the BASE-II
cohort as well (Köhncke et al., 2021).
Venkatesh et al., did, however, report positive associations between older adults’ measure
of hindered-diffusivity and memory performance, arguing that neurite density indexes (NDI)
may serve as a better memory-specific diffusion measure than fractional anisotropy (Venkatesh
et al., 2020). In a follow up study they investigated the relationship between iron concentration
and astrocytic activity across the cortex (Venkatesh et al., 2021). They report that iron-specific
astrogliosis is associated with poor memory performance in the hippocampus and directly relates
RUNNING HEAD: AGING AND LC DIFFUSIVITY
112
to changes in diffusivity (Venkatesh et al., 2021). Interestingly, investigating CA1 and CA3
subregions of the hippocampus revealed higher fractional anisotropy in older adults, relative to
younger adults, and a negative correlation with RAVLT performance (Radhakrishnan et al.,
2022). These study may suggest that the relationship between gray matter diffusivity and
memory may be different than white matter.
Though NDI and fractional anisotropy were reported to be strong predictors of RAVLT
performance, Radhakrishnan et al., note that age may be driving these effects (Radhakrishnan et
al., 2022). We continue this discussion in the summary of Chapter 2. These observations suggest
more research is needed to better understand the relationship between diffusion and memory
performance in gray matter regions of the brain. In our second study we investigate associations
between RAVLT delay scores and diffusivity in the LC and its ascending noradrenergic bundle.
2. Main Finding 2: Fractional anisotropy in the LC displays atypical non-linear
trajectories with aging, compared with the ascending noradrenergic bundle

In Chapter 2, we turned to the NKI dataset, a public dataset with over 300 participants
ranging from 6 to 80 years old. Here, we aimed to replicate previous findings of atypical
diffusivity in the LC between young and older adults and employ generalized additive modeling
(GAM) to explore the non-linear relationships between age and fractional anisotropy. Given the
role of the LC and the ascending noradrenergic bundle in tau pathology, we also measured axial
diffusivity which is commonly utilized in neurodegenerative studies to measure axonal injury
(Bendlin et al., 2012; Brickman et al., 2012; Mac Donald et al., 2007; Mielke et al., 2012;
Molinuevo et al., 2014). The NKI dataset also included 169 participants with two follow-up
visits from their initial baseline visit, allowing us to also investigate changes in diffusivity across
an average of three years. For the longitudinal cohort we expanded our GAM analyses to
RUNNING HEAD: AGING AND LC DIFFUSIVITY
113
generalized additive mixed modeling, which shares some features with non-linear mixed
modeling (Pedersen et al., 2019).
Specifically, GAM allows us to study non-linear relationships and using restricted
maximum likelihood to determine the best lambda value for model selection (Camara et al.,
2019; Pedersen et al., 2019; Sui et al., 2012; Sørensen, Brandmaier, et al., 2021; Sørensen,
Walhovd, et al., 2021). Therefore, we do not have to choose a polynomial degree and some
biases in our model are reduced (Pedersen et al., 2019). Utilizing GAM in neuroimaging studies
has recently become popular with advanced statistical packages in R specific to multivariate
analyses (Sørensen, Brandmaier, et al., 2021; Sørensen, Walhovd, et al., 2021). We also included
Rey Auditory Verbal Learning (RAVLT) delay scores in our models because previous studies
have identified associations between LC diffusion and RAVLT delay scores (Langley et al.,
2020). LC MRI-contrast values have also been reported to correlate with memory performance
(Bachman et al., 2021; Dahl et al., 2021; Dahl, Mather, Werkle-Bergner, et al., 2020).
First, we replicated our findings from the BASE-II and LEMON datasets, observing
lower fractional anisotropy in the LC of young adults relative to older adults. However, we did
not observe any significant difference in ANB fractional anisotropy between older and younger
adults. Second, we reported significant non-linear associations with fractional anisotropy and
axial diffusivity in the LC, ANB, and frontopontine tract across the lifespan in the cross-sectional
cohort (see Chapter 2). Specifically, we observed reductions in fractional anisotropy in the ANB
when compared to the locus coeruleus, and an increase in axial diffusivity in the ANB when
compared to the locus coeruleus.
We observed a positive association between fractional anisotropy in the ANB and
RAVLT delay scores for older adults, compared with young adults. However, we did not observe
RUNNING HEAD: AGING AND LC DIFFUSIVITY
114
any other significant associations between diffusion or fractional anisotropy within the LC,
ANB, or the frontopontine tract and RAVLT delay scores (see Chapter 2 supplementary tables).
Lastly, with the longitudinal cohort, we did not observe any significant changes in LC fractional
anisotropy or LC axial diffusivity over time. Both young and older adults, however, exhibited
significant age-expected non-linear changes in ANB axial diffusivity and fractional anisotropy
across baseline and follow-up visits. For the longitudinal changes in the frontopontine tract we
only exhibited significant associations with fractional anisotropy and axial diffusivity in the
young adults, compared with older adults. We did not observe significant relationships between
RAVLT and any of our ROIs in the longitudinal cohort after correcting for multiple comparisons
(see supplementary tables in Chapter 2).
Nonlinear trajectories of fractional anisotropy and diffusion across the brain have been
reported in the literature (Beck et al., 2021; Kiely et al., 2022; Lebel & Beaulieu, 2011; A. N.
Voineskos et al., 2012). Observations of increasing fractional anisotropy are common from early
childhood until midlife, followed by reductions into late life (Kiely et al., 2022; Lebel &
Beaulieu, 2011). Accordingly, mean, radial, and axial diffusivities have been reported to
decrease from early childhood into midlife, and then increase into late life (Kiely et al., 2022;
Lebel & Beaulieu, 2011). Though we observed some of these characteristics, LC fractional
anisotropy and axial diffusivity were not completely in line with age-expected diffusion
trajectories (Beaudet et al., 2020; Beck et al., 2021; Hasan et al., 2007; Kiely et al., 2022;
Kochunov et al., 2012; Voineskos et al., 2012). Quantifying differences in our GAM smooth
terms of age between the LC and ANB yielded increasing LC fractional anisotropy into late life
in comparison to the ANB (see Chapter 2). Therefore, we hypothesize that microstructural
properties of the LC may affect diffusivity differently, in comparison with the ANB.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
115
Changes in diffusivity in the brains of young adults (5-30 years old) have been associated
with continuing brain development (Lebel & Beaulieu, 2011). It is suggested that observed
changes in diffusivity of young adults is not correlated with brain volume but microstructural
properties of myelin and axonal density (Lebel & Beaulieu, 2011). The specific underlying
neuronal mechanisms that are associated with why diffusivity peaks and declines at specific ages
is still not well understood. A study recently validated fractional anisotropy and radial diffusivity
as a sensitive measure to myelin integrity (Lazari & Lipp, 2021). Specifically, they reported
correlations between fractional anisotropy, radial diffusivity, mean diffusivity, and myelin in a
71-study validation meta-analyses that reviewed microstructural diffusion MR imaging with
myelin histology (Lazari & Lipp, 2021).
Studies have shown that myelin in gray matter regions decreases in a linear fashion,
beginning in early adulthood, and displays an inverted U-shape trajectory in white matter regions
that peaks around 40 years of age (Sherin et al., 2011). Studies have also reported that dendritic
pruning and loss of myelin may be associated with observed reductions in fractional anisotropy
that begin at midlife (Hasan et al., 2007). The speculation that increased microglial and
astrocytic activity may be associated with decreased diffusivity to “counter-balance” the loss in
neuronal and dendritic densities has been proposed by others (Hasan et al., 2007). To understand
the microstructural properties resulting in fractional anisotropy peaking around age 35, studies
propose pairing multiple diffusion techniques, like NODDI and restriction spectrum imaging,
with classical diffusion measurements like fractional anisotropy (Beck et al., 2021).
Regarding the lack of RAVLT associations in our study, one reason may be due to the
healthy cohort, which may not be representative of the typical aging population. Langley et al.,
reported older adults with decreased diffusivity were observed to have worse RAVLT delay
RUNNING HEAD: AGING AND LC DIFFUSIVITY
116
recall scores (Langley et al., 2020; Langley et al., 2021). Langley et al., report that a decrease in
LC diffusivity is likely associated with reduced LC axonal size (Langley et al., 2020). This is in
line with the inverse relationship between mean and radial diffusivity and with fractional
anisotropy that Langley and our lab have reported (Langley et al., 2020; Porat et al., 2022). Other
studies have noted fractional anisotropy to positively correlate with axonal density as well
(Friedrich et al., 2020).
In a follow up study, Langley et al., replicated and expanded their RAVLT and LC
findings and observed higher diffusivity in subregions of the LC, as well as, LC fibers projecting
through the central tegmental tract in older adults, to be associated with better RAVLT
performance (Langley et al., 2021). But the lack of associations in RAVLT delay scores and LC
fractional anisotropy and LC diffusivity from our study as well as young adults from Langley et
al., is somewhat perplexing. One possibility is that young adults are performing too well on the
RAVLT delay and LC diffusivity is not able to capture this relationship; Langley et al., also had
a limited sample size of ~30 participants in each age group which may not be replicable in our
larger 300+ participant dataset (Langley et al., 2021).
If increased LC diffusivity in older adults is associated with better RAVLT delay scores,
then there may be microstructural properties unique to the LC that are at play. Other studies
investigating the associations between hippocampal diffusivity and memory performance, for
example, have observed opposite trends. Lower apparent diffusion coefficients in temporal gray
matter regions are associated with poor memory performances (Sasson et al., 2012). Similarly,
studies report that higher mean diffusivity in the hippocampus was associated with poor verbal
memory performance in adults older than 55 and no associations were observed with fractional
anisotropy (den Heijer, der Lijn, Vernooij, de Groot, Koudstaal, der Lugt, et al., 2012). Poor
RUNNING HEAD: AGING AND LC DIFFUSIVITY
117
verbal memory performance has also been associated with increased mean diffusivity in the
hippocampus (van Norden et al., 2012).
Our results and the available literature offer more questions than answers to the role of
LC diffusion and memory performance between young and older adults. While we expanded the
findings of Chapter 1 and provide further evidence that LC diffusivity and fractional anisotropy
may exhibit atypical age-related trajectories across the lifespan, we are still interested in a
possible mechanism that explains these findings. Specifically, what underlying neuronal
mechanisms may be driving diffusivity differences observed in the LC of young and older
adults? Given the healthy state of the cohort in each data set, in combination with the LC’s role
in sleep-wake transitions, we turned our eyes to sleep deprivation studies as they report
decreased fractional anisotropy in healthy young adults (Elvsåshagen et al., 2015; Rocklage et
al., 2009; Voldsbekk et al., 2021).
3. Main Finding 3: Sleep deprivation may affect fractional anisotropy in the LC
of younger adults more than older adults

In our final Chapter, for Study 3 we turned to the Stockholm Sleepy Brain (SLEEPY)
dataset (Akerstedt, 2016). To our knowledge, SLEEPY is the only publicly available dataset
containing sleep deprived young and older adults with diffusion MRI. In the SLEEPY study,
participants were assigned to both sleep deprived and sleep rested conditions and they completed
an MRI scan during each condition’s visit (Akerstedt, 2016). However, diffusion MRI was only
acquired once, either during the sleep rested or sleep deprived condition, therefore, the study is
restricted to investigating diffusion between, not within, sleep conditions. Given the limited
availability of sleep deprived diffusion MRI datasets, we believe this data may still provide
unique insights in the relationship of LC diffusivity.
RUNNING HEAD: AGING AND LC DIFFUSIVITY
118
Utilizing the same three ROI atlases, we compared diffusivity in the LC, ANB, and
frontopontine tract between young and older adults who were sleep rested or sleep deprived. We
employed an age (young vs older), by ROI (LC, ANB, or frontopontine), by sleep condition
(deprived vs rested) ANOVA. In post-hoc analyses we expanded our investigation with t-tests to
assess within age-group differences in diffusion between sleep deprived and sleep rested young
or older adults. In Chapter 3 we report a trending (p= 0.056) interaction between age, sleep
condition, and ROI. We observed no significant fractional anisotropy or mean and radial
diffusivity differences between sleep rested young and sleep rested older adults (see Chapter 3).
However, we observed higher fractional anisotropy in the LC of sleep deprived older
adults, relative to sleep deprived young adults. In post-hoc analyses we observed that young
adults who were sleep rested displayed higher fractional anisotropy in the LC in comparison to
young adults who were sleep deprived. There were no significant differences in LC fractional
anisotropy between sleep rested and sleep deprived older adults. Notably, we observed
significantly higher ANB fractional anisotropy in young adults, relative to older adults in both
sleep rested and sleep deprived conditions. We did not observe significant differences in ANB
fractional anisotropy within sleep rested and sleep deprived young or older adults.
Thus, the findings in Chapter 3 shift our previous research question from ‘why do older
adults display higher fractional anisotropy relative to younger adults?’ to ‘could sleep
deprivation affect young adult LC fractional anisotropy more than older adults?’ Given the role
of the LC in sleep-wake transitions and inactivity during REM sleep, it is not out of range to
speculate that sleep deprivation may affect diffusion and fractional anisotropy in the LC
(Berridge & Waterhouse, 2003; Hayat et al., 2020; Khanday et al., 2016; Mir & Jha, 2021;
Takahashi et al., 2010). After all, recent studies have reported reductions in fractional anisotropy
RUNNING HEAD: AGING AND LC DIFFUSIVITY
119
in young adults following a single night of sleep deprivation (Elvsåshagen et al., 2015; Khalsa et
al., 2017; Rocklage et al., 2009; Voldsbekk et al., 2021). Studies have also reported reductions in
fractional anisotropy in insomnia patients relative to healthy controls (Bresser et al., 2020; Cai et
al., 2019; Kang et al., 2018; Rostampour et al., 2022; Spiegelhalder et al., 2014; Wu et al., 2020).
So if we consider that LC diffusivity may be affected by sleep deprivation, then the
neuronal underlying mechanisms of the brain’s glymphatic system – where toxic waste is
drained through movement of interstitial fluid while we sleep – may be a relevant target for
future follow-up studies to focus on (Benveniste et al., 2019). Specifically, studies show that LC
noradrenergic activity has direct effect on the glymphatic system’s astrocytes and their aquaporin
channel end feet (AQP4) (Giorgi et al., 2020; Haydon, 2017). AQP4 water channels are critical
to the proper function of the glymphatic system as they assist transporting CSF from perivascular
space to lymphatic vessels outside of the brain (Benveniste et al., 2019; Jessen et al., 2015).
Notably, a recent study reported that sleep deprived wildtype mice were observed with
significantly increased activation of astrocytes, in comparison to controls, but AQP4-deletion
mice displayed both reduced astrocytic activity and increased amyloid and tau burden in
comparison with control mice (Zhang et al., 2020). Sleep deprived wildtype mice displayed
normal amyloid and tau levels in the brain, which was speculated to associate with the increase
in astrocytic activity that may serve as a compensation mechanism for sleep deprivation (Zhang
et al., 2020). Increased astrocytic activity has also been associated with abnormal noradrenaline
release (which occurs during sleep deprivation), and results in reduced axonal volume
(Benveniste et al., 2017; Giorgi et al., 2020; Mendelsohn & Larrick, 2013; Smith et al., 2017).
Under normal waking conditions, the release of noradrenaline can provide anti-
inflammatory responses that reduce astrocytic swelling (Zorec et al., 2017). However, during
RUNNING HEAD: AGING AND LC DIFFUSIVITY
120
sleep deprivation, increased LC noradrenaline release can result in pro-inflammatory responses
that directly modulate NA-K ATPase and astrocytic activity (Mallick & Singh, 2011). Sleep
deprivation also results in LC cell density reduction and subsequent astrocytic swelling can
overtake entire synapses (Bellesi & Dringenberg, 2019; Petit & Magistretti, 2016). Some studies
have also reported direct associations between reduced fractional anisotropy and increased
astrocytic swelling (Blumenfeld-Katzir et al., 2011; Budde et al., 2011; Lee et al., 2021; Singh et
al., 2016).
One study observed in acute injury of swine a reduction in fractional anisotropy which
was directly associated with reduced AQP4 channel activity and increased astrocytic swelling
(Lee et al., 2021). Researchers have also observed increased astrocytic activity in young adults
relative to older adults (Bekar et al., 2008; Florian et al., 2011; Schiweck et al., 2018; Vardjan et
al., 2017). And age-related reductions in LC activity may result in reduced astrocytic activity, as
well as, greater astrocytic atrophy (Zorec et al., 2017). In general, sleep studies have observed
significantly worse outcomes in emotion regulation, reaction time, and cognitive performance in
sleep deprived young adults when compared to older adults (Duffy et al., 2009; Greer et al.,
2013; Nilsonne et al., 2017; Talbot et al., 2010; Zitting et al., 2018). Age differences in LC and
astrocytic activity during sleep deprivation may therefore influence LC diffusivity differently in
young and older adults.
Specifically, it is speculated that sleep deprivation in young adults result in greater
astrocytic activation. We consider the possibility that during sleep deprivation, increased
noradrenaline released from the LC may result in greater astrocytic reactivity, which
consequently reduces axonal volume and fractional anisotropy. Given age-related declines in LC
and astrocytic activity, older adults may not be as susceptible to changes in diffusivity from a
RUNNING HEAD: AGING AND LC DIFFUSIVITY
121
single night of sleep deprivation. Ultimately, more research is necessary. The results of Chapter
3 provide interesting questions, and we encourage future follow-up studies to investigate the
associations between LC diffusion and fractional anisotropy in sleep deprived young and older
adults.
4. Concluding Remarks

In summary, this dissertation provides evidence to suggest that diffusivity in the LC does
not adhere to normal age-related trajectories, in comparison with its ascending noradrenergic
bundle and the frontopontine tract. We offer speculation regarding the underlying neuronal
mechanisms that, with proper replication, may explain our atypical LC diffusivity findings in the
context of sleep deprivation. Advanced neuroimaging techniques like NODDI and proton
magnetic resonance spectroscopy to assess astrocytic metabolism in high resolution data should
be implemented in future studies as well as recommended by others (Beck et al., 2021).
A possible follow up study is sleep depriving humans from REM sleep to observe the
effects of sleep deprivation on LC diffusivity and associated memory performances in young and
older adults. Mice studies could be combined with electrophysiology to investigate the effects of
A1 (pro-inflammatory) and A2 (anti-inflammatory) astrocyte reactivity during sleep deprivation
in relation to LC noradrenaline release and diffusivity. Overall, our understanding of the
microstructural properties of the LC are limited. This dissertation reports on differences in LC
diffusion between young and older adults that indicate that the LC’s microstructural properties
that do not adhere to typical age-expected changes. Given the role of the LC in sleep, arousal,
mood, and Alzheimer’s tau pathology, the LC is an important target of interest to study in-vivo in
humans. With further research, diffusion tensor imaging may provide clinical biomarkers that
RUNNING HEAD: AGING AND LC DIFFUSIVITY
122
detect microstructural changes linked with LC-specific sleep disturbance or Alzheimer’s related
pathology.

RUNNING HEAD: AGING AND LC DIFFUSIVITY
123
References

Abe, Y., Tsurugizawa, T., & Le Bihan, D. (2017). Water diffusion closely reveals neural activity
status in rat brain loci affected by anesthesia. PLoS Biol, 15(4), e2001494.
https://doi.org/10.1371/journal.pbio.2001494
Acosta-Cabronero, J., Williams, G. B., Pengas, G., & Nestor, P. J. (2010). Absolute diffusivities
define the landscape of white matter degeneration in Alzheimer’s disease. Brain, 133(2),
529–539. https://doi.org/10.1093/brain/awp257
Ahlmann-Eltze, C. (2019). ggsignif: Significance Brackets for 'ggplot2'. R package version
0.5.0. In.
Akerstedt, G. N. T. d. O. Å. T. S. L. J. L. N. M. S. A. F. K. F. L. (2016). The Stockholm Sleepy
Brain Study: Effects of Sleep Deprivation on Cognitive and Emotional Processing in
Young and Old. https://doi.org/10.18112/openneuro.ds000201.v1.0.3
Assaf, Y. (2019). Imaging laminar structures in the gray matter with diffusion MRI. Neuroimage,
197, 677-688. https://doi.org/10.1016/j.neuroimage.2017.12.096
Aston-Jones, G., & Waterhouse, B. (2016). Locus coeruleus: From global projection system to
adaptive regulation of behavior. Brain Research, 1645, 75-78.
https://doi.org/https://doi.org/10.1016/j.brainres.2016.03.001
Aust, F., & Barth, M. (2020). papaja: Prepare reproducible APA journal articles with R
Markdown. In (0.1.0.9997 ed.).
Avants, B. B., Epstein, C. L., Grossman, M., & Gee, J. C. (2008). Symmetric diffeomorphic
image registration with cross-correlation: evaluating automated labeling of elderly and
neurodegenerative brain. Med Image Anal, 12(1), 26-41.
https://doi.org/10.1016/j.media.2007.06.004
Babaeeghazvini, P., Rueda-Delgado, L. M., Gooijers, J., Swinnen, S. P., & Daffertshofer, A.
(2021). Brain Structural and Functional Connectivity: A Review of Combined Works of
Diffusion Magnetic Resonance Imaging and Electro-Encephalography. Frontiers in
Human Neuroscience, 15.
https://www.frontiersin.org/article/10.3389/fnhum.2021.721206
Babayan, A., Erbey, M., Kumral, D., Reinelt, J. D., Reiter, A. M. F., Röbbig, J., . . . Villringer,
A. (2019). A mind-brain-body dataset of MRI, EEG, cognition, emotion, and peripheral
physiology in young and old adults. Scientific Data, 6(1), 180308.
https://doi.org/10.1038/sdata.2018.308
Bachman, S. L., Dahl, M. J., Werkle-Bergner, M., Düzel, S., Forlim, C. G., Lindenberger, U., . . .
Mather, M. (2021). Locus coeruleus MRI contrast is associated with cortical thickness in
older adults. Neurobiol Aging, 100, 72-82.
https://doi.org/10.1016/j.neurobiolaging.2020.12.019
Bates, D., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting Linear Mixed-Effects Models
Using lme4. Journal of Statistical Software, 67, 1–48.
https://doi.org/10.18637/jss.v067.i01
RUNNING HEAD: AGING AND LC DIFFUSIVITY
124
Beaudet, G., Tsuchida, A., Petit, L., Tzourio, C., Caspers, S., Schreiber, J., . . . Mazoyer, B.
(2020). Age-Related Changes of Peak Width Skeletonized Mean Diffusivity (PSMD)
Across the Adult Lifespan: A Multi-Cohort Study. Front Psychiatry, 11, 342.
https://doi.org/10.3389/fpsyt.2020.00342
Beck, D., de Lange, A.-M. G., Maximov, I. I., Richard, G., Andreassen, O. A., Nordvik, J. E., &
Westlye, L. T. (2021). White matter microstructure across the adult lifespan: A mixed
longitudinal and cross-sectional study using advanced diffusion models and brain-age
prediction. NeuroImage, 224, 117441.
https://doi.org/https://doi.org/10.1016/j.neuroimage.2020.117441
Bekar, L. K., He, W., & Nedergaard, M. (2008). Locus Coeruleus α-Adrenergic–Mediated
Activation of Cortical Astrocytes In Vivo. Cerebral Cortex, 18(12), 2789-2795.
https://doi.org/10.1093/cercor/bhn040
Bekar, L. K., Wei, H. S., & Nedergaard, M. (2012). The locus coeruleus-norepinephrine network
optimizes coupling of cerebral blood volume with oxygen demand. J Cereb Blood Flow
Metab, 32(12), 2135-2145. https://doi.org/10.1038/jcbfm.2012.115
Bellesi, M., & Dringenberg, H. C. (2019). Chapter 36 - The Effects of Sleep Loss on Brain
Functioning. In Handbook of Behavioral Neuroscience (Vol. 30, pp. 545-556). Elsevier.
https://doi.org/https://doi.org/10.1016/B978-0-12-813743-7.00036-0
Bellesi, M., de Vivo, L., Chini, M., Gilli, F., Tononi, G., & Cirelli, C. (2017). Sleep Loss
Promotes Astrocytic Phagocytosis and Microglial Activation in Mouse Cerebral Cortex. J
Neurosci, 37(21), 5263-5273. https://doi.org/10.1523/JNEUROSCI.3981-16.2017
Bellesi, M., de Vivo, L., Tononi, G., & Cirelli, C. (2015). Effects of sleep and wake on
astrocytes: clues from molecular and ultrastructural studies. BMC Biology, 13(1), 66.
https://doi.org/10.1186/s12915-015-0176-7
Bender, A. R., Brandmaier, A. M., Düzel, S., Keresztes, A., Pasternak, O., Lindenberger, U., &
Kühn, S. (2019). Hippocampal Subfields and Limbic White Matter Jointly Predict
Learning Rate in Older Adults. Cereb Cortex. https://doi.org/10.1093/cercor/bhz252
Bendlin, B. B., Carlsson, C. M., Johnson, S. C., Zetterberg, H., Blennow, K., Willette, A. A., . . .
Sager, M. A. (2012). CSF T-Tau/Aβ42 predicts white matter microstructure in healthy
adults at risk for Alzheimer's disease. PloS one, 7(6), e37720-e37720.
https://doi.org/10.1371/journal.pone.0037720
Benveniste, H., Lee, H., Ding, F., Sun, Q., Al-Bizri, E., Makaryus, R., . . . Lu, H. (2017).
Anesthesia with Dexmedetomidine and Low-dose Isoflurane Increases Solute Transport
via the Glymphatic Pathway in Rat Brain When Compared with High-dose Isoflurane.
Anesthesiology, 127(6), 976-988. https://doi.org/10.1097/ALN.0000000000001888
Benveniste, H., Liu, X., Koundal, S., Sanggaard, S., Lee, H., & Wardlaw, J. (2019). The
Glymphatic System and Waste Clearance with Brain Aging: A Review. Gerontology,
65(2), 106-119. https://doi.org/10.1159/000490349
Beppu, T., Inoue, T., Shibata, Y., Kurose, A., Arai, H., Ogasawara, K., . . . Kabasawa, H. (2003).
Measurement of fractional anisotropy using diffusion tensor MRI in supratentorial
astrocytic tumors. Journal of Neuro-Oncology, 63(2), 109-116.
https://doi.org/10.1023/A:1023977520909
RUNNING HEAD: AGING AND LC DIFFUSIVITY
125
Berridge, C. W., & Waterhouse, B. D. (2003). The locus coeruleus-noradrenergic system:
modulation of behavioral state and state-dependent cognitive processes. Brain Res Brain
Res Rev, 42(1), 33-84. https://doi.org/10.1016/s0165-0173(03)00143-7
Bertram, L., Böckenhoff, A., Demuth, I., Düzel, S., Eckardt, R., Li, S. C., . . . Steinhagen-
Thiessen, E. (2014). Cohort profile: The Berlin Aging Study II (BASE-II). Int J
Epidemiol, 43(3), 703-712. https://doi.org/10.1093/ije/dyt018
Bhagat, Y. A., & Beaulieu, C. (2004). Diffusion anisotropy in subcortical white matter and
cortical gray matter: Changes with aging and the role of CSF-suppression. Journal of
Magnetic Resonance Imaging, 20(2), 216-227. https://doi.org/10.1002/jmri.20102
Bijanki, K. R., Matsui, J. T., Mayberg, H. S., Magnotta, V. A., Arndt, S., Johnson, H. J.,
Nopoulos, P., Paradiso, S., McCormick, L. M., Fiedorowicz, J. G., Epping, E. A., &
Moser, D. J. (2015). Depressive symptoms related to low fractional anisotropy of white
matter underlying the right ventral anterior cingulate in older adults with atherosclerotic
vascular disease. Frontiers in Human Neuroscience, 9, 408.
https://doi.org/10.3389/fnhum.2015.00408
Blumenfeld-Katzir, T., Pasternak, O., Dagan, M., & Assaf, Y. (2011). Diffusion MRI of
structural brain plasticity induced by a learning and memory task. PLoS One, 6(6),
e20678. https://doi.org/10.1371/journal.pone.0020678
Bozzali, M., Falini, A., Franceschi, M., Cercignani, M., Zuffi, M., Scotti, G., . . . Filippi, M.
(2002). White matter damage in Alzheimer's disease assessed in vivo using diffusion
tensor magnetic resonance imaging. J Neurol Neurosurg Psychiatry, 72(6), 742-746.
https://doi.org/10.1136/jnnp.72.6.742
Braak, H., & Del Trecidi, K. (2015). Neuroanatomy and pathology of sporadic Alzheimer's
disease. Adv Anat Embryol Cell Biol, 215, 1-162.
Braak, H., Thal, D. R., Ghebremedhin, E., & Del Tredici, K. (2011). Stages of the pathologic
process in Alzheimer disease: age categories from 1 to 100 years. J Neuropathol Exp
Neurol, 70(11), 960-969. https://doi.org/10.1097/NEN.0b013e318232a379
Bresser, T., Foster-Dingley, J. C., Wassing, R., Leerssen, J., Ramautar, J. R., Stoffers, D., . . .
van Someren, E. J. W. (2020). Consistent altered internal capsule white matter
microstructure in insomnia disorder. Sleep, 43(8). https://doi.org/10.1093/sleep/zsaa031
Brickman, A. M., Meier, I. B., Korgaonkar, M. S., Provenzano, F. A., Grieve, S. M., Siedlecki,
K. L., . . . Zimmerman, M. E. (2012). Testing the white matter retrogenesis hypothesis of
cognitive aging. Neurobiology of Aging, 33(8), 1699-1715.
https://doi.org/https://doi.org/10.1016/j.neurobiolaging.2011.06.001
Budde, M. D., Janes, L., Gold, E., Turtzo, L. C., & Frank, J. A. (2011). The contribution of
gliosis to diffusion tensor anisotropy and tractography following traumatic brain injury:
validation in the rat using Fourier analysis of stained tissue sections. Brain, 134(Pt 8),
2248-2260. https://doi.org/10.1093/brain/awr161
Bussian, T. J., Aziz, A., Meyer, C. F., Swenson, B. L., van Deursen, J. M., & Baker, D. J. (2018).
Clearance of senescent glial cells prevents tau-dependent pathology and cognitive
decline. Nature, 562(7728), 578-582. https://doi.org/10.1038/s41586-018-0543-y
RUNNING HEAD: AGING AND LC DIFFUSIVITY
126
Cai, W., Zhao, M., Liu, J., Liu, B., Yu, D., & Yuan, K. (2019). Right arcuate fasciculus and
superior longitudinal fasciculus abnormalities in primary insomnia. Brain Imaging
Behav, 13(6), 1746-1755. https://doi.org/10.1007/s11682-019-00160-1
Camara, E. M., Costa de Azevedo, M. C., Franco, T. P., & Araújo, F. G. (2019). Hierarchical
partitioning of fish diversity and scale-dependent environmental effects in tropical coastal
ecosystems. Mar Environ Res, 148, 26-38.
https://doi.org/10.1016/j.marenvres.2019.05.006
Cascio, C., Gribbin, M., Gouttard, S., Smith, R. G., Jomier, M., Field, S., Graves, M., Hazlett, H.
C., Muller, K., Gerig, G., & Piven, J. (2013). Fractional anisotropy distributions in 2- to
6-year-old children with autism. Journal of Intellectual Disability Research : JIDR,
57(11), 1037–1049. https://doi.org/10.1111/j.1365-2788.2012.01599.x
Cedernaes, J., Osorio, R. S., Varga, A. W., Kam, K., Schiöth, H. B., & Benedict, C. (2017).
Candidate mechanisms underlying the association between sleep-wake disruptions and
Alzheimer's disease. Sleep Medicine Reviews, 31, 102-111.
https://doi.org/https://doi.org/10.1016/j.smrv.2016.02.002
Chalermpalanupap, T., Weinshenker, D., & Rorabaugh, J. M. (2017). Down but Not Out: The
Consequences of Pretangle Tau in the Locus Coeruleus. Neural Plasticity, 2017,
7829507. https://doi.org/10.1155/2017/7829507
Chang, E. H., Argyelan, M., Aggarwal, M., Chandon, T.-S. S., Karlsgodt, K. H., Mori, S., &
Malhotra, A. K. (2017). The role of myelination in measures of white matter integrity:
Combination of diffusion tensor imaging and two-photon microscopy of CLARITY intact
brains. NeuroImage, 147, 253–261. https://doi.org/10.1016/j.neuroimage.2016.11.068
Chary, K., Narvaez, O., Salo, R. A., San Martín Molina, I., Tohka, J., Aggarwal, M., . . . Sierra,
A. (2021). Microstructural Tissue Changes in a Rat Model of Mild Traumatic Brain
Injury. Front Neurosci, 15, 746214. https://doi.org/10.3389/fnins.2021.746214
Chen, X., Huddleston, D. E., Langley, J., Ahn, S., Barnum, C. J., Factor, S. A., . . . Hu, X.
(2014). Simultaneous imaging of locus coeruleus and substantia nigra with a quantitative
neuromelanin MRI approach. Magn Reson Imaging, 32(10), 1301-1306.
https://doi.org/10.1016/j.mri.2014.07.003
Chu, W. T., Wang, W.-e., Zaborszky, L., Golde, T. E., DeKosky, S., Duara, R., . . . Vaillancourt,
D. E. (2021). Association of Cognitive Impairment With Free Water in the Nucleus
Basalis of Meynert and Locus Coeruleus to Transentorhinal Cortex Tract. Neurology,
10.1212/WNL.0000000000013206. https://doi.org/10.1212/WNL.0000000000013206
Clewett, D. V., Huang, R., Velasco, R., Lee, T. H., & Mather, M. (2018). Locus Coeruleus
Activity Strengthens Prioritized Memories Under Arousal. In J Neurosci (2018/01/04 ed.,
Vol. 38, Issue 6, pp. 1558–1574).
Clewett, D. V., Lee, T.-H., Greening, S., Ponzio, A., Margalit, E., & Mather, M. (2016).
Neuromelanin marks the spot: identifying a locus coeruleus biomarker of cognitive
reserve in healthy aging. Neurobiology of aging, 37, 117-126.
https://doi.org/10.1016/j.neurobiolaging.2015.09.019
RUNNING HEAD: AGING AND LC DIFFUSIVITY
127
Clewett, D., Bachman, S., & Mather, M. (2014). Age-related reduced prefrontal-amygdala
structural connectivity is associated with lower trait anxiety. Neuropsychology, 28(4),
631-642. https://doi.org/10.1037/neu0000060
Core Team, R. (2019). R: A language and environment for statistical computing. R Foundation
for Statistical Computing. https://www.R-project.org/
Dahl, M. J., Mather, M., Düzel, S., Bodammer, N. C., Lindenberger, U., Kühn, S., & Werkle-
Bergner, M. (2019). Rostral locus coeruleus integrity is associated with better memory
performance in older adults. Nat Hum Behav, 3(11), 1203-1214.
https://doi.org/10.1038/s41562-019-0715-2
Dahl, M. J., Mather, M., Sander, M. C., & Werkle-Bergner, M. (2020). Noradrenergic
Responsiveness Supports Selective Attention across the Adult Lifespan. The Journal of
Neuroscience, 40(22), 4372. https://doi.org/10.1523/JNEUROSCI.0398-19.2020
Dahl, M. J., Mather, M., Werkle-Bergner, M., Kennedy, B. L., Guzman, S., Hurth, K., . . .
Ringman, J. M. (2022). Locus coeruleus integrity is related to tau burden and memory
loss in autosomal-dominant Alzheimer's disease. Neurobiology of Aging, 112, 39-54.
https://doi.org/https://doi.org/10.1016/j.neurobiolaging.2021.11.006
Dahl, M. J., Mather, M., Werkle-Bergner, M., Kennedy, B. L., Qiao, Y., Shi, Y., & Ringman, J.
M. (2020). Lower MR-indexed locus coeruleus integrity in autosomal-dominant
Alzheimer’s disease is related to cortical tau burden and memory deficits. medRxiv,
2020.2011.2016.20232561. https://doi.org/10.1101/2020.11.16.20232561
Delius, M, J. A., Düzel , S., Gerstorf , D., & Lindenberger , U. (2015). Berlin Aging Studies
(BASE and BASE-II). In Pachana (Ed.), Encyclopedia of geropsychology (pp. 386–395):
Springer.
den Heijer, T., der Lijn, F., Vernooij, M. W., de Groot, M., Koudstaal, P. J., van der Lugt, A., . . .
Breteler, M. M. (2012). Structural and diffusion MRI measures of the hippocampus and
memory performance. Neuroimage, 63(4), 1782-1789.
https://doi.org/10.1016/j.neuroimage.2012.08.067
Dinis, J., & Bragança, M. (2018). Quality of Sleep and Depression in College Students: A
Systematic Review. Sleep Science, 11(4), 290–301. https://doi.org/10.5935/1984-
0063.20180045
Dinov, I. D., Van Horn, J. D., Lozev, K. M., Magsipoc, R., Petrosyan, P., Liu, Z., . . . Toga, A.
W. (2009). Efficient, Distributed and Interactive Neuroimaging Data Analysis Using the
LONI Pipeline. Frontiers in neuroinformatics, 3, 22-22.
https://doi.org/10.3389/neuro.11.022.2009
DiNuzzo, M., & Nedergaard, M. (2017). Brain energetics during the sleep–wake cycle. Current
Opinion in Neurobiology, 47, 65-72.
https://doi.org/https://doi.org/10.1016/j.conb.2017.09.010
Douaud, G., Menke, R. A., Gass, A., Monsch, A. U., Rao, A., Whitcher, B., . . . Smith, S. (2013).
Brain microstructure reveals early abnormalities more than two years prior to clinical
progression from mild cognitive impairment to Alzheimer's disease. J Neurosci, 33(5),
2147-2155. https://doi.org/10.1523/JNEUROSCI.4437-12.2013
RUNNING HEAD: AGING AND LC DIFFUSIVITY
128
Duffy, J. F., Willson, H. J., Wang, W., & Czeisler, C. A. (2009). Healthy older adults better
tolerate sleep deprivation than young adults. Journal of the American Geriatrics Society,
57(7), 1245-1251. https://doi.org/10.1111/j.1532-5415.2009.02303.x
Dzierzewski, J. M., Donovan, E. K., Kay, D. B., Sannes, T. S., & Bradbrook, K. E. (2020). Sleep
Inconsistency and Markers of Inflammation. Frontiers in Neurology, 11.
https://www.frontiersin.org/article/10.3389/fneur.2020.01042
Edlow, B. L., McNab, J. A., Witzel, T., & Kinney, H. C. (2016). The Structural Connectome of
the Human Central Homeostatic Network. Brain Connect, 6(3), 187-200.
https://doi.org/10.1089/brain.2015.0378
Edlow, B. L., Takahashi, E., Wu, O., Benner, T., Dai, G., Bu, L., Grant, P. E., Greer, D. M.,
Greenberg, S. M., Kinney, H. C., & Folkerth, R. D. (2012). Neuroanatomic connectivity
of the human ascending arousal system critical to consciousness and its disorders. In J
Neuropathol Exp Neurol (Vol. 71, Issue 6, pp. 531–546).
Ehrenberg, A. J., Suemoto, C. K., França Resende, E. d. P., Petersen, C., Leite, R. E. P.,
Rodriguez, R. D., . . . Grinberg, L. T. (2018). Neuropathologic Correlates of Psychiatric
Symptoms in Alzheimer’s Disease. Journal of Alzheimer's Disease, 66, 115-126.
https://doi.org/10.3233/JAD-180688
Elsayed, H., Faroni, A., Ashraf, M. R., Osuji, J., Wunderley, L., Zhang, L., . . . Reid, A. J.
(2020). Development and Characterisation of an. Front Bioeng Biotechnol, 8, 784.
https://doi.org/10.3389/fbioe.2020.00784
Elvsåshagen, T., Norbom, L. B., Pedersen, P. Ø., Quraishi, S. H., Bjørnerud, A., Malt, U. F.,
Groote, I. R., & Westlye, L. T. (2015). Widespread Changes in White Matter
Microstructure after a Day of Waking and Sleep Deprivation. PLOS ONE, 10(5),
e0127351. https://doi.org/10.1371/journal.pone.0127351
Ethell, D. W. (2014). Disruption of cerebrospinal fluid flow through the olfactory system may
contribute to Alzheimer’s disease pathogenesis. In J Alzheimers Dis (Vol. 41, Issue 4, pp.
1021–1030).
Florian, C., Vecsey, C. G., Halassa, M. M., Haydon, P. G., & Abel, T. (2011). Astrocyte-derived
adenosine and A1 receptor activity contribute to sleep loss-induced deficits in
hippocampal synaptic plasticity and memory in mice. J Neurosci, 31(19), 6956-6962.
https://doi.org/10.1523/JNEUROSCI.5761-10.2011
Friedrich, P., Fraenz, C., Schlüter, C., Ocklenburg, S., Mädler, B., Güntürkün, O., & Genç, E.
(2020). The Relationship Between Axon Density, Myelination, and Fractional Anisotropy
in the Human Corpus Callosum. Cereb Cortex, 30(4), 2042-2056.
https://doi.org/10.1093/cercor/bhz221
Giorgi, F. S., Galgani, A., Puglisi-Allegra, S., Limanaqi, F., Busceti, C. L., & Fornai, F. (2020).
Locus Coeruleus and neurovascular unit: From its role in physiology to its potential role
in Alzheimer's disease pathogenesis. J Neurosci Res, 98(12), 2406-2434.
https://doi.org/10.1002/jnr.24718
Goldstein, A. N., & Walker, M. P. (2014). The role of sleep in emotional brain function. Annual
review of clinical psychology, 10, 679-708. https://doi.org/10.1146/annurev-clinpsy-
032813-153716
RUNNING HEAD: AGING AND LC DIFFUSIVITY
129
Gong, L., Shi, M., Wang, J., Xu, R., Yu, S., Liu, D., . . . Xi, C. (2021). The Abnormal Functional
Connectivity in the Locus Coeruleus-Norepinephrine System Associated With Anxiety
Symptom in Chronic Insomnia Disorder. Frontiers in neuroscience, 15, 678465-678465.
https://doi.org/10.3389/fnins.2021.678465
Grambaite, R., Stenset, V., Reinvang, I., Walhovd, K. B., Fjell, A. M., & Fladby, T. (2010).
White matter diffusivity predicts memory in patients with subjective and mild cognitive
impairment and normal CSF total tau levels. Journal of the International
Neuropsychological Society, 16(1), 58-69. https://doi.org/DOI:
10.1017/S1355617709990932
Grandner, M. A., Alfonso-Miller, P., Fernandez-Mendoza, J., Shetty, S., Shenoy, S., & Combs,
D. (2016). Sleep: important considerations for the prevention of cardiovascular disease.
Current opinion in cardiology, 31(5), 551-565.
https://doi.org/10.1097/HCO.0000000000000324
Greer, S. M., Goldstein, A. N., & Walker, M. P. (2013). The impact of sleep deprivation on food
desire in the human brain. Nat Commun, 4, 2259. https://doi.org/10.1038/ncomms3259
Grumbach, P., Opel, N., Martin, S., Meinert, S., Leehr, E. J., Redlich, R., Enneking, V.,
Goltermann, J., Baune, B. T., Dannlowski, U., & Repple, J. (2020). Sleep duration is
associated with white matter microstructure and cognitive performance in healthy adults.
Human Brain Mapping, 41(15), 4397–4405. https://doi.org/10.1002/hbm.25132
Gutchess, A., Thomas, A. K., & Mather, M. (2020). How Arousal-Related Neurotransmitter
Systems Compensate for Age-Related Decline. In The Cambridge Handbook of
Cognitive Aging: A Life Course Perspective (pp. 101-120). Cambridge University Press.
https://doi.org/DOI: 10.1017/9781108552684.007
Halassa, M. M., Florian, C., Fellin, T., Munoz, J. R., Lee, S. Y., Abel, T., . . . Frank, M. G.
(2009). Astrocytic modulation of sleep homeostasis and cognitive consequences of sleep
loss. Neuron, 61(2), 213-219. https://doi.org/10.1016/j.neuron.2008.11.024
Han, S. D., Boyle, P. A., Arfanakis, K., Fleischman, D., Yu, L., James, B. D., & Bennett, D. A.
(2016). Financial literacy is associated with white matter integrity in old age.
NeuroImage, 130, 223-229. https://doi.org/10.1016/j.neuroimage.2016.02.030
Harley, C. W., Walling, S. G., Yuan, Q., & Martin, G. M. (2021). The ‘a, b, c’s of pretangle tau
and their relation to aging and the risk of Alzheimer’s Disease. Seminars in Cell &
Developmental Biology, 116, 125-134.
https://doi.org/https://doi.org/10.1016/j.semcdb.2020.12.010
Harris, N. G., Verley, D. R., Gutman, B. A., & Sutton, R. L. (2016). Bi-directional changes in
fractional anisotropy after experiment TBI: Disorganization and reorganization?
Neuroimage, 133, 129-143. https://doi.org/10.1016/j.neuroimage.2016.03.012
Hasan, K. M., Sankar, A., Halphen, C., Kramer, L. A., Brandt, M. E., Juranek, J., . . . Ewing-
Cobbs, L. (2007). Development and organization of the human brain tissue compartments
across the lifespan using diffusion tensor imaging. Neuroreport, 18(16), 1735-1739.
https://doi.org/10.1097/WNR.0b013e3282f0d40c
RUNNING HEAD: AGING AND LC DIFFUSIVITY
130
Hayat, H., Regev, N., Matosevich, N., Sales, A., Paredes-Rodriguez, E., Krom, A. J., . . . Nir, Y.
(2020). Locus coeruleus norepinephrine activity mediates sensory-evoked awakenings
from sleep. Sci Adv, 6(15), eaaz4232. https://doi.org/10.1126/sciadv.aaz4232
Haydon, P. G. (2017). Astrocytes and the modulation of sleep. Current Opinion in Neurobiology,
44, 28-33. https://doi.org/https://doi.org/10.1016/j.conb.2017.02.008
Henf, J., Grothe, M. J., Brueggen, K., Teipel, S., & Dyrba, M. (2018). Mean diffusivity in
cortical gray matter in Alzheimer’s disease: The importance of partial volume correction.
NeuroImage: Clinical, 17, 579–586. https://doi.org/10.1016/j.nicl.2017.10.005
Holth Jerrah, K., Fritschi Sarah, K., Wang, C., Pedersen Nigel, P., Cirrito John, R., Mahan
Thomas, E., . . . Holtzman David, M. (2019). The sleep-wake cycle regulates brain
interstitial fluid tau in mice and CSF tau in humans. Science, 363(6429), 880-884.
https://doi.org/10.1126/science.aav2546
Iram, T., Trudler, D., Kain, D., Kanner, S., Galron, R., Vassar, R., . . . Frenkel, D. (2016).
Astrocytes from old Alzheimer's disease mice are impaired in Aβ uptake and in
neuroprotection. Neurobiology of Disease, 96, 84-94.
https://doi.org/https://doi.org/10.1016/j.nbd.2016.08.001
Irwin, M. R., Olmstead, R., & Carroll, J. E. (2016). Sleep Disturbance, Sleep Duration, and
Inflammation: A Systematic Review and Meta-Analysis of Cohort Studies and
Experimental Sleep Deprivation. Biol Psychiatry, 80(1), 40-52.
https://doi.org/10.1016/j.biopsych.2015.05.014
Iwasa, K., Tasaki, I., & Gibbons, R. C. (1980). Swelling of nerve fibers associated with action
potentials. Science, 210(4467), 338-339. https://doi.org/10.1126/science.7423196
Jacobs, H. I. L., Müller-Ehrenberg, L., Priovoulos, N., & Roebroeck, A. (2018). Curvilinear
locus coeruleus functional connectivity trajectories over the adult lifespan: A 7T MRI
study. Neurobiology of Aging, 69, 167–176.
https://doi.org/10.1016/j.neurobiolaging.2018.05.021
Jacobs, H. I., van Boxtel, M. P., Gronenschild, E. H., Uylings, H. B., Jolles, J., & Verhey, F. R.
(2013). Decreased gray matter diffusivity: a potential early Alzheimer's disease
biomarker? Alzheimers Dement, 9(1), 93-97. https://doi.org/10.1016/j.jalz.2011.11.004
Jessen, N. A., Munk, A. S. F., Lundgaard, I., & Nedergaard, M. (2015). The Glymphatic System:
A Beginner's Guide. Neurochemical research, 40(12), 2583-2599.
https://doi.org/10.1007/s11064-015-1581-6
Ju, K.-Y., Lee, J. W., Im, G. H., Lee, S., Pyo, J., Park, S. B., Lee, J. H., & Lee, J.-K. (2013). Bio-
inspired, melanin-like nanoparticles as a highly efficient contrast agent for T1-weighted
magnetic resonance imaging. Biomacromolecules, 14(10), 3491–3497.
https://doi.org/10.1021/bm4008138
Ju, Y.-E. S., Ooms, S. J., Sutphen, C., Macauley, S. L., Zangrilli, M. A., Jerome, G., . . .
Holtzman, D. M. (2017). Slow wave sleep disruption increases cerebrospinal fluid
amyloid-β levels. Brain, 140(8), 2104-2111. https://doi.org/10.1093/brain/awx148

RUNNING HEAD: AGING AND LC DIFFUSIVITY
131
Kang, J. M. K., Joo, S. W. J., Son, Y. D. S., Kim, H. K., Ko, K. P. K., Lee, J. S. L., & Kang, S.
G. K. (2018). Low white-matter integrity between the left thalamus and inferior frontal
gyrus in patients with insomnia disorder. J Psychiatry Neurosci, 43(6), 366-374.
https://doi.org/10.1503/jpn.170195
Kang, S. S., Ahn, E. H., & Ye, K. (2020). Delta-secretase cleavage of Tau mediates its pathology
and propagation in Alzheimer's disease. Exp Mol Med, 52(8), 1275-1287.
https://doi.org/10.1038/s12276-020-00494-7
Kantarci, K. (2014). Fractional Anisotropy of the Fornix and Hippocampal Atrophy in
Alzheimer’s Disease. Frontiers in Aging Neuroscience, 6, 316.
Kantarci, K., Murray, M. E., Schwarz, C. G., Reid, R. I., Przybelski, S. A., Lesnick, T., . . .
Dickson, D. W. (2017). White-matter integrity on DTI and the pathologic staging of
Alzheimer's disease. Neurobiology of aging, 56, 172-179.
https://doi.org/10.1016/j.neurobiolaging.2017.04.024
Kantarci, K., Murray, M. E., Schwarz, C. G., Reid, R. I., Przybelski, S. A., Lesnick, T., . . .
Dickson, D. W. (2017). White-matter integrity on DTI and the pathologic staging of
Alzheimer's disease. Neurobiology of aging, 56, 172-179.
https://doi.org/10.1016/j.neurobiolaging.2017.04.024
Kantarci, K., Petersen, R. C., Boeve, B. F., Knopman, D. S., Weigand, S. D., O'Brien, P. C., . . .
Jack, C. R. (2005). DWI predicts future progression to Alzheimer disease in amnestic
mild cognitive impairment. Neurology, 64(5), 902-904.
https://doi.org/10.1212/01.WNL.0000153076.46126.E9
Kassambara, A. rstatix: Pipe-Friendly Framework for Basic Statistical Tests. In (0.7.0 ed.).
Kato, H., Okuno, T., Isohashi, K., Koda, T., Shimizu, M., Mochizuki, H., . . . Hatazawa, J.
(2021). Astrocyte metabolism in multiple sclerosis investigated by 1-C-11 acetate PET. J
Cereb Blood Flow Metab, 41(2), 369-379. https://doi.org/10.1177/0271678X20911469
Keren, N. I., Lozar, C. T., Harris, K. C., Morgan, P. S., & Eckert, M. A. (2009). In vivo mapping
of the human locus coeruleus. NeuroImage, 47(4), 1261–1267.
Keren, N. I., Taheri, S., Vazey, E. M., Morgan, P. S., Granholm, A.-C. E., Aston-Jones, G. S., &
Eckert, M. A. (2015). Histologic validation of locus coeruleus MRI contrast in post-
mortem tissue. NeuroImage, 113, 235–245.
Khalsa, S., Hale, J. R., Goldstone, A., Wilson, R. S., Mayhew, S. D., Bagary, M., & Bagshaw, A.
P. (2017). Habitual sleep durations and subjective sleep quality predict white matter
differences in the human brain. Neurobiol Sleep Circadian Rhythms, 3, 17-25.
https://doi.org/10.1016/j.nbscr.2017.03.001
Khanday, M. A., Somarajan, B. I., Mehta, R., & Mallick, B. N. (2016). Noradrenaline from
Locus Coeruleus Neurons Acts on Pedunculo-Pontine Neurons to Prevent REM Sleep
and Induces Its Loss-Associated Effects in Rats. eNeuro, 3(6).
https://doi.org/10.1523/ENEURO.0108-16.2016

RUNNING HEAD: AGING AND LC DIFFUSIVITY
132
Kiely, M., Triebswetter, C., Cortina, L. E., Gong, Z., Alsameen, M. H., Spencer, R. G., &
Bouhrara, M. (2022). Insights into human cerebral white matter maturation and
degeneration across the adult lifespan. NeuroImage, 247, 118727.
https://doi.org/https://doi.org/10.1016/j.neuroimage.2021.118727
Kochunov, P., Williamson, D. E., Lancaster, J., Fox, P., Cornell, J., Blangero, J., & Glahn, D. C.
(2012). Fractional anisotropy of water diffusion in cerebral white matter across the
lifespan. Neurobiol Aging, 33(1), 9-20.
https://doi.org/10.1016/j.neurobiolaging.2010.01.014
Kochunov, P., Williamson, D. E., Lancaster, J., Fox, P., Cornell, J., Blangero, J., & Glahn, D. C.
(2012). Fractional anisotropy of water diffusion in cerebral white matter across the
lifespan. Neurobiol Aging, 33(1), 9-20.
https://doi.org/10.1016/j.neurobiolaging.2010.01.014
Köhncke, Y., Düzel, S., Sander, M. C., Lindenberger, U., Kühn, S., & Brandmaier, A. M.
(2021). Hippocampal and Parahippocampal Gray Matter Structural Integrity Assessed by
Multimodal Imaging Is Associated with Episodic Memory in Old Age. Cereb Cortex,
31(3), 1464-1477. https://doi.org/10.1093/cercor/bhaa287
Kulesh, A., Drobakha, V., Kuklina, E., Nekrasova, I., & Shestakov, V. (2018). Cytokine
Response, Tract-Specific Fractional Anisotropy, and Brain Morphometry in Post-Stroke
Cognitive Impairment. Journal of Stroke and Cerebrovascular Diseases, 27(7), 1752–
1759. https://doi.org/10.1016/j.jstrokecerebrovasdis.2018.02.004
Langley, J., Hussain, S., Flores, J. J., Bennett, I. J., & Hu, X. (2020). Characterization of age-
related microstructural changes in locus coeruleus and substantia nigra pars compacta.
Neurobiol Aging, 87, 89-97. https://doi.org/10.1016/j.neurobiolaging.2019.11.016
Langley, J., Hussain, S., Huddleston, D. E., Bennett, I. J., & Hu, X. P. (2021). Impact of Locus
Coeruleus and Its Projections on Memory and Aging. Brain Connectivity.
https://doi.org/10.1089/brain.2020.0947
Lavoie, C. J., Zeidler, M. R., & Martin, J. L. (2018). Sleep and aging. Sleep Science and
Practice, 2(1), 3. https://doi.org/10.1186/s41606-018-0021-3
Lawrence, K. E., Nabulsi, L., Santhalingam, V., Abaryan, Z., Villalon-Reina, J. E., Nir, T. M., . .
. Thompson, P. M. (2021). Age and sex effects on advanced white matter microstructure
measures in 15,628 older adults: A UK biobank study. Brain Imaging and Behavior,
15(6), 2813-2823. https://doi.org/10.1007/s11682-021-00548-y
Lazari, A., & Lipp, I. (2021). Can MRI measure myelin? Systematic review, qualitative
assessment, and meta-analysis of studies validating microstructural imaging with myelin
histology. Neuroimage, 230, 117744. https://doi.org/10.1016/j.neuroimage.2021.117744
Le Bihan, D., Urayama, S., Aso, T., Hanakawa, T., & Fukuyama, H. (2006). Direct and fast
detection of neuronal activation in the human brain with diffusion MRI. Proc Natl Acad
Sci U S A, 103(21), 8263-8268. https://doi.org/10.1073/pnas.0600644103
Lebel, C., & Beaulieu, C. (2011). Longitudinal development of human brain wiring continues
from childhood into adulthood. J Neurosci, 31(30), 10937-10947.
https://doi.org/10.1523/JNEUROSCI.5302-10.2011
RUNNING HEAD: AGING AND LC DIFFUSIVITY
133
Lee, D. H., Park, J. W., Park, S. H., & Hong, C. (2015). Have You Ever Seen the Impact of
Crossing Fiber in DTI?: Demonstration of the Corticospinal Tract Pathway. PLoS One,
10(7), e0112045. https://doi.org/10.1371/journal.pone.0112045
Lee, J. K., Liu, D., Jiang, D., Kulikowicz, E., Tekes, A., Liu, P., . . . Martin, L. J. (2021).
Fractional anisotropy from diffusion tensor imaging correlates with acute astrocyte and
myelin swelling in neonatal swine models of excitotoxic and hypoxic-ischemic brain
injury. J Comp Neurol, 529(10), 2750-2770. https://doi.org/10.1002/cne.25121
Lee, T.-H., Greening, S. G., Ueno, T., Clewett, D., Ponzio, A., Sakaki, M., & Mather, M. (2018).
Arousal increases neural gain via the locus coeruleus–noradrenaline system in younger
adults but not in older adults. Nature Human Behaviour, 2(5), 356-366.
https://doi.org/10.1038/s41562-018-0344-1
Lenth, R. V. (2021). emmeans: Estimated Marginal Means, aka Least-Squares Means. In (1.5.5-
1 ed.).
Li, X.-Y., Tang, Z.-C., Sun, Y., Tian, J., Liu, Z.-Y., & Han, Y. (2016). White matter
degeneration in subjective cognitive decline: a diffusion tensor imaging study.
Oncotarget, 7(34), 54405-54414. https://doi.org/10.18632/oncotarget.10091
Liu, K. Y., Acosta-Cabronero, J., Cardenas-Blanco, A., Loane, C., Berry, A. J., Betts, M. J., . . .
Cam-CAN. (2019). In vivo visualization of age-related differences in the locus coeruleus.
Neurobiol Aging, 74, 101-111. https://doi.org/10.1016/j.neurobiolaging.2018.10.014
Liu, K. Y., Kievit, R. A., Tsvetanov, K. A., Betts, M. J., Düzel, E., Rowe, J. B., . . . Cam-CAN.
(2020). Noradrenergic-dependent functions are associated with age-related locus
coeruleus signal intensity differences. Nat Commun, 11(1), 1712.
https://doi.org/10.1038/s41467-020-15410-w
Liu, K. Y., Marijatta, F., Hämmerer, D., Acosta-Cabronero, J., Düzel, E., & Howard, R. J.
(2017). Magnetic resonance imaging of the human locus coeruleus: A systematic review.
Neurosci Biobehav Rev, 83, 325-355. https://doi.org/10.1016/j.neubiorev.2017.10.023
Loizou, L. A. (1969). Projections of the nucleus locus coeruleus in the albino rat. Brain
Research, 15(2), 563-566. https://doi.org/https://doi.org/10.1016/0006-8993(69)90185-1
Loughlin, S. E., Foote, S. L., & Fallon, J. H. (1982). Locus coeruleus projections to cortex:
Topography, morphology and collateralization. Brain Research Bulletin, 9(1), 287-294.
https://doi.org/https://doi.org/10.1016/0361-9230(82)90142-3
Lucey, B. P., Wisch, J., Boerwinkle, A. H., Landsness, E. C., Toedebusch, C. D., McLeland, J.
S., Butt, O. H., Hassenstab, J., Morris, J. C., Ances, B. M., & Holtzman, D. M. (2021).
Sleep and longitudinal cognitive performance in preclinical and early symptomatic
Alzheimer’s disease. Brain, 144(9), 2852–2862. https://doi.org/10.1093/brain/awab272
Mac Donald, C. L., Dikranian, K., Song, S. K., Bayly, P. V., Holtzman, D. M., & Brody, D. L.
(2007). Detection of traumatic axonal injury with diffusion tensor imaging in a mouse
model of traumatic brain injury. Experimental neurology, 205(1), 116-131.
https://doi.org/10.1016/j.expneurol.2007.01.035

RUNNING HEAD: AGING AND LC DIFFUSIVITY
134
Makowski, D., Ben-Shachar, M. S., Patil, I., & Lüdecke, D. (2020). Estimation of Model-Based
Predictions, Contrasts and Means. https://github.com/easystats/modelbased
Mallick, B. N., & Singh, A. (2011). REM sleep loss increases brain excitability: Role of
noradrenalin and its mechanism of action. Sleep Medicine Reviews, 15(3), 165-178.
https://doi.org/https://doi.org/10.1016/j.smrv.2010.11.001
Mander, B. A., Winer, J. R., & Walker, M. P. (2017). Sleep and Human Aging. Neuron, 94(1),
19-36. https://doi.org/10.1016/j.neuron.2017.02.004
Mander, B. A., Winer, J. R., Jagust, W. J., & Walker, M. P. (2016). Sleep: A Novel Mechanistic
Pathway, Biomarker, and Treatment Target in the Pathology of Alzheimer's Disease?
Trends in Neurosciences, 39(8), 552-566.
https://doi.org/https://doi.org/10.1016/j.tins.2016.05.002
Mather, M. (2020). The locus coeruleus norepinephrine system role in cognition and how it
changes with aging (D. Poeppel, G. Mangung, & M. Gazzaniga, Eds.). MIT Press.
Mather, M. (2021). Noradrenaline in the aging brain: Promoting cognitive reserve or accelerating
Alzheimer's disease? Seminars in Cell & Developmental Biology, 116, 108-124.
https://doi.org/https://doi.org/10.1016/j.semcdb.2021.05.013
Mather, M., & Harley, C. W. (2016). The Locus Coeruleus: Essential for Maintaining Cognitive
Function and the Aging Brain. Trends in Cognitive Sciences, 20(3), 214-226.
https://doi.org/https://doi.org/10.1016/j.tics.2016.01.001
McGregor, R., & Siegel, J. M. (2010). Illuminating the locus coeruleus: control of posture and
arousal. Nature Neuroscience, 13, 1448.
McLean, M. W., Hooker, G., Staicu, A.-M., Scheipl, F., & Ruppert, D. (2014). Functional
Generalized Additive Models. Journal of Computational and Graphical Statistics, 23(1),
249–269. https://doi.org/10.1080/10618600.2012.729985
Medina, & Gaviria. (2008). Diffusion tensor imaging investigations in Alzheimer's disease: the
resurgence of white matter compromise in the cortical dysfunction of the aging brain.
Neuropsychiatric disease and treatment, 4(4), 737-742.
Meinert, S., Repple, J., Nenadic, I., Krug, A., Jansen, A., Grotegerd, D., Förster, K., Enneking,
V., Dohm, K., Schmitt, S., Stein, F., Brosch, K., Meller, T., Redlich, R., Böhnlein, J.,
Sindermann, L., Goltermann, J., Leehr, E. J., Opel, N., … Dannlowski, U. (2019).
Reduced fractional anisotropy in depressed patients due to childhood maltreatment rather
than diagnosis. Neuropsychopharmacology, 44(12), 2065–2072.
https://doi.org/10.1038/s41386-019-0472-y
Mendelsohn, A. R., & Larrick, J. W. (2013). Sleep facilitates clearance of metabolites from the
brain: glymphatic function in aging and neurodegenerative diseases. Rejuvenation Res,
16(6), 518-523. https://doi.org/10.1089/rej.2013.1530
Merenstein, J. L., Corrada, M. M., Kawas, C. H., & Bennett, I. J. (2021). Age affects white
matter microstructure and episodic memory across the older adult lifespan. Neurobiology
of Aging, 106, 282–291. https://doi.org/10.1016/j.neurobiolaging.2021.06.021
Mielke, M. M., Okonkwo, O. C., Oishi, K., Mori, S., Tighe, S., Miller, M. I., . . . Lyketsos, C. G.
(2012). Fornix integrity and hippocampal volume predict memory decline and
RUNNING HEAD: AGING AND LC DIFFUSIVITY
135
progression to Alzheimer’s disease. Alzheimer's & Dementia, 8(2), 105-113.
https://doi.org/https://doi.org/10.1016/j.jalz.2011.05.2416
Mir, F. A., & Jha, S. K. (2021). Locus Coeruleus Acid-Sensing Ion Channels Modulate Sleep–
Wakefulness and State Transition from NREM to REM Sleep in the Rat. Neuroscience
Bulletin, 37(5), 684-700. https://doi.org/10.1007/s12264-020-00625-0
Molinuevo, J. L., Ripolles, P., Simó, M., Lladó, A., Olives, J., Balasa, M., . . . Rami, L. (2014).
White matter changes in preclinical Alzheimer's disease: a magnetic resonance imaging-
diffusion tensor imaging study on cognitively normal older people with positive amyloid
β protein 42 levels. Neurobiology of Aging, 35(12), 2671-2680.
https://doi.org/https://doi.org/10.1016/j.neurobiolaging.2014.05.027
Moradi, E., Hallikainen, I., Hänninen, T., & Tohka, J. (2016). Rey’s Auditory Verbal Learning
Test scores can be predicted from whole brain MRI in Alzheimer’s disease. NeuroImage :
Clinical, 13, 415–427. https://doi.org/10.1016/j.nicl.2016.12.011
Morris, L. S., McCall, J. G., Charney, D. S., & Murrough, J. W. (2020). The role of the locus
coeruleus in the generation of pathological anxiety. Brain Neurosci Adv, 4,
2398212820930321. https://doi.org/10.1177/2398212820930321
Müller, M. J., Greverus, D., Dellani, P. R., Weibrich, C., Wille, P. R., Scheurich, A., . . .
Fellgiebel, A. (2005). Functional implications of hippocampal volume and diffusivity in
mild cognitive impairment. Neuroimage, 28(4), 1033-1042.
https://doi.org/10.1016/j.neuroimage.2005.06.029
Mullington, J. M., Simpson, N. S., Meier-Ewert, H. K., & Haack, M. (2010). Sleep loss and
inflammation. Best Practice & Research Clinical Endocrinology & Metabolism, 24(5),
775–784. https://doi.org/10.1016/j.beem.2010.08.014
Muncy, N. M., Kimbler, A., Hedges-Muncy, A. M., McMakin, D. L., & Mattfeld, A. T. (2022).
General additive models address statistical issues in diffusion MRI: An example with
clinically anxious adolescents. NeuroImage: Clinical, 33, 102937.
https://doi.org/10.1016/j.nicl.2022.102937
Nakamoto, K., Aizawa, F., Kinoshita, M., Koyama, Y., & Tokuyama, S. (2017). Astrocyte
Activation in Locus Coeruleus Is Involved in Neuropathic Pain Exacerbation Mediated
by Maternal Separation and Social Isolation Stress. Frontiers in pharmacology, 8, 401-
401. https://doi.org/10.3389/fphar.2017.00401
Ni, J., Chen, S., Liu, J., Huang, G., Shen, T., & Chen, X. (2010). Regional diffusion changes of
cerebral grey matter during normal aging—A fluid-inversion prepared diffusion imaging
study. European Journal of Radiology, 75(2), 134–138.
https://doi.org/10.1016/j.ejrad.2009.04.028
Nilsonne, G., Tamm, S., Schwarz, J., Almeida, R., Fischer, H., Kecklund, G., . . . Akerstedt, T.
(2017). Intrinsic brain connectivity after partial sleep deprivation in young and older
adults: results from the Stockholm Sleepy Brain study. Sci Rep, 7(1), 9422.
https://doi.org/10.1038/s41598-017-09744-7

RUNNING HEAD: AGING AND LC DIFFUSIVITY
136
Nooner, K. B., Colcombe, S. J., Tobe, R. H., Mennes, M., Benedict, M. M., Moreno, A. L.,
Panek, L. J., Brown, S., Zavitz, S. T., Li, Q., Sikka, S., Gutman, D., Bangaru, S.,
Schlachter, R. T., Kamiel, S. M., Anwar, A. R., Hinz, C. M., Kaplan, M. S., Rachlin, A.
B., … Milham, M. P. (2012). The NKI-Rockland Sample: A Model for Accelerating the
Pace of Discovery Science in Psychiatry. Frontiers in Neuroscience, 6, 152.
https://doi.org/10.3389/fnins.2012.00152
Nunes, D., Gil, R., & Shemesh, N. (2021). A rapid-onset diffusion functional MRI signal reflects
neuromorphological coupling dynamics. Neuroimage, 231, 117862.
https://doi.org/10.1016/j.neuroimage.2021.117862
Oouchi, H., Yamada, K., Sakai, K., Kizu, O., Kubota, T., Ito, H., & Nishimura, T. (2007).
Diffusion Anisotropy Measurement of Brain White Matter Is Affected by Voxel Size:
Underestimation Occurs in Areas with Crossing Fibers. American Journal of
Neuroradiology, 28(6), 1102. https://doi.org/10.3174/ajnr.A0488
Pal, D., Trivedi, R., Saksena, S., Yadav, A., Kumar, M., Pandey, C. M., . . . Gupta, R. K. (2011).
Quantification of age- and gender-related changes in diffusion tensor imaging indices in
deep grey matter of the normal human brain. J Clin Neurosci, 18(2), 193-196.
https://doi.org/10.1016/j.jocn.2010.05.033
Paukert, M., Agarwal, A., Cha, J., Doze, Van A., Kang, Jin U., & Bergles, Dwight E. (2014).
Norepinephrine Controls Astroglial Responsiveness to Local Circuit Activity. Neuron,
82(6), 1263-1270. https://doi.org/https://doi.org/10.1016/j.neuron.2014.04.038
Pedersen, E. J., Miller, D. L., Simpson, G. L., & Ross, N. (2019). Hierarchical generalized
additive models in ecology: an introduction with mgcv. PeerJ, 7, e6876.
https://doi.org/10.7717/peerj.6876
Petit, J. M., & Magistretti, P. J. (2016). Regulation of neuron–astrocyte metabolic coupling
across the sleep–wake cycle. Neuroscience, 323, 135-156.
https://doi.org/https://doi.org/10.1016/j.neuroscience.2015.12.007
Pfefferbaum, A., & Sullivan, E. V. (2003). Increased brain white matter diffusivity in normal
adult aging: relationship to anisotropy and partial voluming. Magn Reson Med, 49(5),
953-961. https://doi.org/10.1002/mrm.10452
Pfefferbaum, A., Adalsteinsson, E., Rohlfing, T., & Sullivan, E. V. (2010). Diffusion tensor
imaging of deep gray matter brain structures: Effects of age and iron concentration.
Neurobiology of Aging, 31(3), 482–493.
https://doi.org/10.1016/j.neurobiolaging.2008.04.013
Popov, A., Brazhe, A., Denisov, P., Sutyagina, O., Li, L., Lazareva, N., . . . Semyanov, A.
(2021). Astrocyte dystrophy in ageing brain parallels impaired synaptic plasticity. Aging
Cell, 20(3), e13334. https://doi.org/https://doi.org/10.1111/acel.13334
Porat, S., Sibilia, F., Yoon, J., Shi, Y., Dahl, M. J., Werkle-Bergner, M., . . . Mather, M. (2022).
Age differences in diffusivity in the locus coeruleus and its ascending noradrenergic tract.
NeuroImage, 251, 119022.
https://doi.org/https://doi.org/10.1016/j.neuroimage.2022.119022
RUNNING HEAD: AGING AND LC DIFFUSIVITY
137
Priovoulos, N., van Boxel, S. C. J., Jacobs, H. I. L., Poser, B. A., Uludag, K., Verhey, F. R. J., &
Ivanov, D. (2020). Unraveling the contributions to the neuromelanin-MRI contrast. Brain
Structure and Function, 225(9), 2757-2774. https://doi.org/10.1007/s00429-020-02153-z
Quattrini, G., Pini, L., Galazzo, I. B., Pizzini, F. B., & Pievani, M. (2021). Microstructural
alterations of the locus coeruleus-entorhinal cortex pathway in Alzheimer’s disease and
behavioral variant frontotemporal dementia. Alzheimer’s & Dementia, 17(S3), e057743.
https://doi.org/10.1002/alz.057743
Radhakrishnan, H., Bennett, I. J., & Stark, C. E. (2022). Higher-order multi-shell diffusion
measures complement tensor metrics and volume in gray matter when predicting age and
cognition. Neuroimage, 253, 119063. https://doi.org/10.1016/j.neuroimage.2022.119063
Raichle, M. E., Hartman, B. K., Eichling, J. O., & Sharpe, L. G. (1975). Central noradrenergic
regulation of cerebral blood flow and vascular permeability. Proc Natl Acad Sci U S A,
72(9), 3726-3730. https://doi.org/10.1073/pnas.72.9.3726
Roalf, D. R., de la Garza, A. G., Rosen, A., Calkins, M. E., Moore, T. M., Quarmley, M.,
Ruparel, K., Xia, C. H., Rupert, P. E., Satterthwaite, T. D., Shinohara, R. T., Elliott, M.
A., Gur, R. C., & Gur, R. E. (2020). Alterations in white matter microstructure in
individuals at persistent risk for psychosis. Molecular Psychiatry, 25(10), 2441–2454.
https://doi.org/10.1038/s41380-019-0360-1
Rocklage, M., Williams, V., Pacheco, J., & Schnyer, D. M. (2009). White Matter Differences
Predict Cognitive Vulnerability to Sleep Deprivation. Sleep, 32(8), 1100-1103.
https://doi.org/10.1093/sleep/32.8.1100
Rose, N. L., Yang, H., Turner, S. D., & Simpson, G. L. (2012). An assessment of the
mechanisms for the transfer of lead and mercury from atmospherically contaminated
organic soils to lake sediments with particular reference to Scotland, UK. Geochimica et
Cosmochimica Acta, 82, 113–135. https://doi.org/10.1016/j.gca.2010.12.026
Rose, S. E., Janke PhD, A. L., & Chalk, J. B. (2008). Gray and white matter changes in
Alzheimer's disease: A diffusion tensor imaging study. Journal of Magnetic Resonance
Imaging, 27(1), 20-26. https://doi.org/https://doi.org/10.1002/jmri.21231
Rostampour, M., Gharaylou, Z., Rostampour, N., Kaveh, D., Noori, K., Fadaei, R., . . . Zarei, M.
(2022). Asymmetric alterations of white matter integrity in patients with insomnia
disorder. Brain Imaging Behav, 16(1), 389-396. https://doi.org/10.1007/s11682-021-
00512-w
Rotshenker, S. (2011). Wallerian degeneration: the innate-immune response to traumatic nerve
injury. J Neuroinflammation, 8, 109. https://doi.org/10.1186/1742-2094-8-109
Ryu, S. Y., Lim, E. Y., Na, S., Shim, Y. S., Cho, J. H., Yoon, B., . . . Yang, D. W. (2017).
Hippocampal and entorhinal structures in subjective memory impairment: a combined
MRI volumetric and DTI study. International Psychogeriatrics, 29(5), 785-792.
https://doi.org/DOI: 10.1017/S1041610216002349

RUNNING HEAD: AGING AND LC DIFFUSIVITY
138
Salminen, L. E., Conturo, T. E., Laidlaw, D. H., Cabeen, R. P., Akbudak, E., Lane, E. M., Heaps,
J. M., Bolzenius, J. D., Baker, L. M., Cooley, S., Scott, S., Cagle, L. M., Phillips, S., &
Paul, R. H. (2016). Regional age differences in gray matter diffusivity among healthy
older adults. Brain Imaging and Behavior, 10(1), 203–211.
https://doi.org/10.1007/s11682-015-9383-7
Sara, S. J. (2009). The locus coeruleus and noradrenergic modulation of cognition. Nature
Reviews Neuroscience, 10, 211.
Sasson, E., Doniger, G. M., Pasternak, O., Tarrasch, R., & Assaf, Y. (2012). Structural correlates
of cognitive domains in normal aging with diffusion tensor imaging. Brain Struct Funct,
217(2), 503-515. https://doi.org/10.1007/s00429-011-0344-7
Schiweck, J., Eickholt, B. J., & Murk, K. (2018). Important Shapeshifter: Mechanisms Allowing
Astrocytes to Respond to the Changing Nervous System During Development, Injury and
Disease. Front Cell Neurosci, 12, 261. https://doi.org/10.3389/fncel.2018.00261
Sexton, C. E., Mackay, C. E., Lonie, J. A., Bastin, M. E., Terrière, E., O'Carroll, R. E., &
Ebmeier, K. P. (2010). MRI correlates of episodic memory in Alzheimer's disease, mild
cognitive impairment, and healthy aging. Psychiatry Research: Neuroimaging, 184(1),
57-62. https://doi.org/https://doi.org/10.1016/j.pscychresns.2010.07.005
SH, T., J, F., JW, C., YY, W., & CCC, C. (2012). Age-related Changes of White Matter in the
Elderly Population Measured by Diffusion Tensor Imaging. In OMICS J Radiology: Vol.
1:107.
Shattuck, D. W., & Leahy, R. M. (2002). BrainSuite: An automated cortical surface
identification tool. In Med Image Anal (Vol. 6, Issue 2, pp. 129–142).
Sherin, J. E., Bartzokis, G., Masoro, E. J., & Austad, S. N. (2011). Chapter 15 - Human Brain
Myelination Trajectories Across the Life Span: Implications for CNS Function and
Dysfunction. In Handbook of the Biology of Aging (Seventh Edition) (pp. 333-346).
Academic Press. https://doi.org/https://doi.org/10.1016/B978-0-12-378638-8.00015-4
Sherpa, A. D., Aoki, C., Hrabetova, S., Vardjan, N., & Zorec, R. (2017). Chapter 12 -
Noradrenaline Drives Structural Changes in Astrocytes and Brain Extracellular Space. In
Noradrenergic Signaling and Astroglia (pp. 241-255). Academic Press.
https://doi.org/https://doi.org/10.1016/B978-0-12-805088-0.00012-8
Sibilia, F., Kehoe, E. G., Farrell, D., Kerskens, C., O'Neill, D., McNulty, J. P., . . . Bokde, A. L.
W. (2017). Aging-Related Microstructural Alterations Along the Length of the Cingulum
Bundle. Brain Connect, 7(6), 366-372. https://doi.org/10.1089/brain.2017.0493
Simpson, G. L. (2022). gratia: Graceful ggplot-Based Graphics and Other Functions
for GAMs Fitted using mgcv. https://gavinsimpson.github.io/gratia/
Singh, K., Trivedi, R., Devi, M. M., Tripathi, R. P., & Khushu, S. (2016). Longitudinal changes
in the DTI measures, anti-GFAP expression and levels of serum inflammatory cytokines
following mild traumatic brain injury. Exp Neurol, 275 Pt 3, 427-435.
https://doi.org/10.1016/j.expneurol.2015.07.016

RUNNING HEAD: AGING AND LC DIFFUSIVITY
139
Sintini, I., Schwarz, C. G., Martin, P. R., Graff‐Radford, J., Machulda, M. M., Senjem, M. L.,
Reid, R. I., Spychalla, A. J., Drubach, D. A., Lowe, V. J., Jack, C. R., Josephs, K. A., &
Whitwell, J. L. (2018). Regional multimodal relationships between tau, hypometabolism,
atrophy, and fractional anisotropy in atypical Alzheimer’s disease. Human Brain
Mapping, 40(5), 1618–1631. https://doi.org/10.1002/hbm.24473
Sjoberg, D. D., Curry, M., Hannum, M., Whiting, K., & Zabor, E. C. (2021). gtsummary:
Presentation-Ready Data Summary and Analytic Result Tables. In (1.3.7 ed.).
Smith, A. J., Yao, X., Dix, J. A., Jin, B. J., & Verkman, A. S. (2017). Test of the 'glymphatic'
hypothesis demonstrates diffusive and aquaporin-4-independent solute transport in rodent
brain parenchyma. Elife, 6. https://doi.org/10.7554/eLife.27679
Smith, S. M., Jenkinson, M., Woolrich, M. W., Beckmann, C. F., Behrens, T. E., Johansen-Berg,
H., . . . Matthews, P. M. (2004). Advances in functional and structural MR image analysis
and implementation as FSL. Neuroimage, 23 Suppl 1, S208-219.
https://doi.org/10.1016/j.neuroimage.2004.07.051
Soares, J. M., Marques, P., Alves, V., & Sousa, N. (2013). A hitchhiker's guide to diffusion
tensor imaging. Frontiers in neuroscience, 7, 31-31.
https://doi.org/10.3389/fnins.2013.00031
Solar, K. G., Treit, S., & Beaulieu, C. (2021). High resolution diffusion tensor imaging of the
hippocampus across the healthy lifespan. Hippocampus, 31(12), 1271-1284.
https://doi.org/10.1002/hipo.23388
Solders, S. K., Galinsky, V. L., Clark, A. L., Sorg, S. F., Weigand, A. J., Bondi, M. W., & Frank,
L. R. (2022). Diffusion MRI tractography of the locus coeruleus-transentorhinal cortex
connections using GO-ESP. Magnetic Resonance in Medicine, 87(4), 1816-1831.
https://doi.org/https://doi.org/10.1002/mrm.29088
Song, I., Neal, J., & Lee, T.-H. (2021). Age-Related Intrinsic Functional Connectivity Changes
of Locus Coeruleus from Childhood to Older Adults. Brain Sciences, 11(11), 1485.
https://doi.org/10.3390/brainsci11111485
Sørensen, Ø., Brandmaier, A. M., Macià, D., Ebmeier, K., Ghisletta, P., Kievit, R. A., . . . Fjell,
A. (2021). Meta-analysis of generalized additive models in neuroimaging studies.
Neuroimage, 224, 117416. https://doi.org/10.1016/j.neuroimage.2020.117416
Sørensen, Ø., Walhovd, K. B., & Fjell, A. M. (2021). A recipe for accurate estimation of lifespan
brain trajectories, distinguishing longitudinal and cohort effects. Neuroimage, 226,
117596. https://doi.org/10.1016/j.neuroimage.2020.117596
Spiegelhalder, K., Regen, W., Prem, M., Baglioni, C., Nissen, C., Feige, B., . . . Riemann, D.
(2014). Reduced anterior internal capsule white matter integrity in primary insomnia.
Hum Brain Mapp, 35(7), 3431-3438. https://doi.org/10.1002/hbm.22412
Stebbins, G. T., & Murphy, C. M. (2009). Diffusion tensor imaging in Alzheimer's disease and
mild cognitive impairment. Behavioural neurology, 21(1), 39-49.
https://doi.org/10.3233/BEN-2009-0234

RUNNING HEAD: AGING AND LC DIFFUSIVITY
140
Stock, G., Rupprecht, U., Stumpf, H., & Schlör, K. H. (1981). Cardiovascular changes during
arousal elicited by stimulation of amygdala, hypothalamus and locus coeruleus. Journal
of the Autonomic Nervous System, 3(2), 503-510.
https://doi.org/https://doi.org/10.1016/0165-1838(81)90083-7
Strain, J. F., Smith, R. X., Beaumont, H., Roe, C. M., Gordon, B. A., Mishra, S., Adeyemo, B.,
Christensen, J. J., Su, Y., Morris, J. C., Benzinger, T. L. S., & Ances, B. M. (2018). Loss
of white matter integrity reflects tau accumulation in Alzheimer disease defined regions.
Neurology, 91(4), e313–e318. https://doi.org/10.1212/WNL.0000000000005864
Sui, J., Adali, T., Yu, Q., Chen, J., & Calhoun, V. D. (2012). A review of multivariate methods
for multimodal fusion of brain imaging data. Journal of Neuroscience Methods, 204(1),
68-81. https://doi.org/https://doi.org/10.1016/j.jneumeth.2011.10.031
Sullivan, E. V., & Pfefferbaum, A. (2006). Diffusion tensor imaging and aging. Neuroscience &
Biobehavioral Reviews, 30(6), 749-761.
https://doi.org/https://doi.org/10.1016/j.neubiorev.2006.06.002
Sulzer, D., Cassidy, C., Horga, G., Kang, U. J., Fahn, S., Casella, L., Pezzoli, G., Langley, J.,
Hu, X. P., Zucca, F. A., Isaias, I. U., & Zecca, L. (2018). Neuromelanin detection by
magnetic resonance imaging (MRI) and its promise as a biomarker for Parkinson’s
disease. Npj Parkinson’s Disease, 4(1), 1–13. https://doi.org/10.1038/s41531-018-0047-3
Sun, W., Tang, Y., Qiao, Y., Ge, X., Mather, M., Ringman, J. M., . . . Initiative, f. A. s. D. N.
(2020). A probabilistic atlas of locus coeruleus pathways to transentorhinal cortex for
connectome imaging in Alzheimer's disease. Neuroimage, 223, 117301.
https://doi.org/10.1016/j.neuroimage.2020.117301
Svoboda, J., & Syková, E. (1991). Extracellular space volume changes in the rat spinal cord
produced by nerve stimulation and peripheral injury. Brain Res, 560(1-2), 216-224.
https://doi.org/10.1016/0006-8993(91)91235-s
Takahashi, K., Kayama, Y., Lin, J. S., & Sakai, K. (2010). Locus coeruleus neuronal activity
during the sleep-waking cycle in mice. Neuroscience, 169(3), 1115-1126.
https://doi.org/10.1016/j.neuroscience.2010.06.009
Talbot, L. S., McGlinchey, E. L., Kaplan, K. A., Dhal, R. E., & Harvey, A. G. (2010). Sleep
deprivation in adolescents and adults: changes in affect. Emotion, 10(6), 831-841.
https://doi.org/10.1037/a0020138
Tang, Y., Sun, W., Toga, A. W., Ringman, J. M., & Shi, Y. (2018). A probabilistic atlas of
human brainstem pathways based on connectome imaging data. NeuroImage, 169, 227-
239. https://doi.org/https://doi.org/10.1016/j.neuroimage.2017.12.042
Team, R. C. (2020). R: A Language and Environment for Statistical Computing. R Foundation
for Statistical Computing. https://www.R-project.org/
Tirosh, N., & Nevo, U. (2013). Neuronal activity significantly reduces water displacement: DWI
of a vital rat spinal cord with no hemodynamic effect. Neuroimage, 76, 98-107.
https://doi.org/10.1016/j.neuroimage.2013.02.065

RUNNING HEAD: AGING AND LC DIFFUSIVITY
141
Tournier, J. D., Smith, R., Raffelt, D., Tabbara, R., Dhollander, T., Pietsch, M., . . . Connelly, A.
(2019). MRtrix3: A fast, flexible and open software framework for medical image
processing and visualisation. Neuroimage, 202, 116137.
https://doi.org/10.1016/j.neuroimage.2019.116137
Trujillo, P., Petersen, K. J., Cronin, M. J., Lin, Y.-C., Kang, H., Donahue, M. J., . . . Claassen, D.
O. (2019). Quantitative magnetization transfer imaging of the human locus coeruleus.
NeuroImage, 200, 191-198.
https://doi.org/https://doi.org/10.1016/j.neuroimage.2019.06.049
van Norden, A. G., de Laat, K. F., Fick, I., van Uden, I. W., van Oudheusden, L. J., Gons, R. A.,
. . . de Leeuw, F. E. (2012). Diffusion tensor imaging of the hippocampus and verbal
memory performance: the RUN DMC study. Hum Brain Mapp, 33(3), 542-551.
https://doi.org/10.1002/hbm.21231
Vanek, J., Prasko, J., Genzor, S., Ociskova, M., Kantor, K., Holubova, M., Slepecky, M.,
Nesnidal, V., Kolek, A., & Sova, M. (2020). Obstructive sleep apnea, depression and
cognitive impairment. Sleep Medicine, 72, 50–58.
https://doi.org/10.1016/j.sleep.2020.03.017
Vardjan, N., Verkhratsky, A., & Zorec, R. (2017). Astrocytic Pathological Calcium Homeostasis
and Impaired Vesicle Trafficking in Neurodegeneration. Int J Mol Sci, 18(2).
https://doi.org/10.3390/ijms18020358
Venkatesh, A., Daugherty, A. M., & Bennett, I. J. (2021). Neuroimaging measures of iron and
gliosis explain memory performance in aging. Hum Brain Mapp, 42(17), 5761-5770.
https://doi.org/10.1002/hbm.25652
Venkatesh, A., Stark, S. M., Stark, C. E. L., & Bennett, I. J. (2020). Age- and memory- related
differences in hippocampal gray matter integrity are better captured by NODDI compared
to single-tensor diffusion imaging. Neurobiol Aging, 96, 12-21.
https://doi.org/10.1016/j.neurobiolaging.2020.08.004
Voineskos, A. N., Rajji, T. K., Lobaugh, N. J., Miranda, D., Shenton, M. E., Kennedy, J. L., . . .
Mulsant, B. H. (2012). Age-related decline in white matter tract integrity and cognitive
performance: a DTI tractography and structural equation modeling study. Neurobiol
Aging, 33(1), 21-34. https://doi.org/10.1016/j.neurobiolaging.2010.02.009
Voldsbekk, I., Groote, I., Zak, N., Roelfs, D., Geier, O., Due-Tønnessen, P., . . . Maximov, I. I.
(2021). Sleep and sleep deprivation differentially alter white matter microstructure: A
mixed model design utilising advanced diffusion modelling. Neuroimage, 226, 117540.
https://doi.org/10.1016/j.neuroimage.2020.117540
Voldsbekk, I., Maximov, I. I., Zak, N., Roelfs, D., Geier, O., Due-Tønnessen, P., . . . Groote, I.
(2020). Evidence for wakefulness-related changes to extracellular space in human brain
white matter from diffusion-weighted MRI. Neuroimage, 212, 116682.
https://doi.org/10.1016/j.neuroimage.2020.116682
Walker, K. A., Power, M. C., Hoogeveen, R. C., Folsom, A. R., Ballantyne, C. M., Knopman, D.
S., Windham, B. G., Selvin, E., Jack, C. R., & Gottesman, R. F. (2017). Midlife Systemic
Inflammation, Late-Life White Matter Integrity, and Cerebral Small Vessel Disease.
Stroke, 48(12), 3196–3202. https://doi.org/10.1161/STROKEAHA.117.018675
RUNNING HEAD: AGING AND LC DIFFUSIVITY
142
Weinshenker, D. (2018). Long Road to Ruin: Noradrenergic Dysfunction in Neurodegenerative
Disease. Trends Neurosci, 41(4), 211-223. https://doi.org/10.1016/j.tins.2018.01.010
Westlye, L. T., Walhovd, K. B., Dale, A. M., Bjørnerud, A., Due-Tønnessen, P., Engvig, A.,
Grydeland, H., Tamnes, C. K., Ostby, Y., & Fjell, A. M. (2010). Life-span changes of the
human brain white matter: Diffusion tensor imaging (DTI) and volumetry. In Cereb
Cortex (2009/12/23 ed., Vol. 20, Issue 9, pp. 2055–2068).
Weston, P. S. J., Simpson, I. J. A., Ryan, N. S., Ourselin, S., & Fox, N. C. (2015). Diffusion
imaging changes in grey matter in Alzheimer’s disease: a potential marker of early
neurodegeneration. Alzheimer's Research & Therapy, 7(1), 47.
https://doi.org/10.1186/s13195-015-0132-3
Wicham, H. (2017). tidyverse: Easily Install and Load the 'Tidyverse'. R package version 1.2.1.
https://CRAN.R-project.org/package=tidyverse
Wickham, H. (2016). ggplot2: Elegant Graphics for Data Analysis. In: Springer-Verlag New
York.
Winter, B., & Wieling, M. (2016). How to analyze linguistic change using mixed models,
Growth Curve Analysis and Generalized Additive Modeling. Journal of Language
Evolution, 1(1), 7–18. https://doi.org/10.1093/jole/lzv003
Wood, S. N. (2011). Fast stable restricted maximum likelihood and marginal likelihood
estimation of semiparametric generalized linear models. Journal of the Royal Statistical
Society: Series B (Statistical Methodology), 73(1), 3–36. https://doi.org/10.1111/j.1467-
9868.2010.00749.x
Wu, Y., Zhuang, Y., & Qi, J. (2020). Explore structural and functional brain changes in insomnia
disorder: A PRISMA-compliant whole brain ALE meta-analysis for multimodal MRI.
Medicine (Baltimore), 99(14), e19151. https://doi.org/10.1097/MD.0000000000019151
Xie, Y. (2021). knitr: A General-Purpose Package for Dynamic Report Generation in R. In (1.31
ed.).
Yi, S. Y., Barnett, B. R., Torres-Velázquez, M., Zhang, Y., Hurley, S. A., Rowley, P. A., . . . Yu,
J. J. (2019). Detecting Microglial Density With Quantitative Multi-Compartment
Diffusion MRI. Front Neurosci, 13, 81. https://doi.org/10.3389/fnins.2019.00081
Zhang, R., Liu, Y., Chen, Y., Li, Q., Marshall, C., Wu, T., . . . Xiao, M. (2020). Aquaporin 4
deletion exacerbates brain impairments in a mouse model of chronic sleep disruption.
CNS Neurosci Ther, 26(2), 228-239. https://doi.org/10.1111/cns.13194
Zitting, K. M., Münch, M. Y., Cain, S. W., Wang, W., Wong, A., Ronda, J. M., . . . Duffy, J. F.
(2018). Young adults are more vulnerable to chronic sleep deficiency and recurrent
circadian disruption than older adults. Sci Rep, 8(1), 11052.
https://doi.org/10.1038/s41598-018-29358-x
Zorec, R., Vardjan, N., & Verkhratsky, A. (2017). Chapter 1 - Locus Coeruleus Noradrenergic
Neurons and Astroglia in Health and Disease. In Noradrenergic Signaling and Astroglia
(pp. 1-24). Academic Press. https://doi.org/https://doi.org/10.1016/B978-0-12-805088-
0.00001-3
RUNNING HEAD: AGING AND LC DIFFUSIVITY
143
Zucca, F. A., Bellei, C., Giannelli, S., Terreni, M. R., Gallorini, M., Rizzio, E., Pezzoli, G.,
Albertini, A., & Zecca, L. (2006). Neuromelanin and iron in human locus coeruleus and
substantia nigra during aging: Consequences for neuronal vulnerability. Journal of Neural
Transmission, 113(6), 757–767. https://doi.org/10.1007/s00702-006-0453-2
Zuurbier, L. A., Nikolova, Y. S., Åhs, F., & Hariri, A. R. (2013). Uncinate fasciculus fractional
anisotropy correlates with typical use of reappraisal in women but not men. Emotion
(Washington, D.C.), 13(3), 385-390. https://doi.org/10.1037/a0031163

Abstract (if available)

Abstract The locus coeruleus (LC) is essential for arousal and attentional behavior. As the brain’s primary source for noradrenaline, the health of the LC is paramount to the central nervous system. Previous studies have identified the LC as the initial site of abnormal tau in the Alzheimer’s disease progression of pathology. However, in-vivo investigations of the LC have been difficult given its small size and deep brainstem location. Recently, neuroimaging techniques like diffusion tensor imaging, have allowed us to study the microstructural integrity of the LC in-vivo in humans. For example, fractional anisotropy is widely used as a measure of structural integrity (higher fractional anisotropy typically indicates healthier axons) and has a strong inverse correlation with axial, radial and mean diffusivities.

The aims of this dissertation were to identify age-related diffusivity of the LC in relation to the LC ascending noradrenergic bundle, a white matter tract that projects axons from the LC and terminates in the entorhinal cortex. Specifically, my goal was to first replicate existing literature reports of increased fractional anisotropy in the LC of older adults, relative to younger adults (an atypical age-related discovery). Then I investigated if these changes in diffusion are observed consistently throughout the lifespan. Given the LC ascending noradrenergic bundle terminates in the entorhinal cortex, I also investigated its relationship to verbal memory. Finally, I attempted to explore possible underlying mechanisms that may be explain age-related diffusivity differences in the LC with a sleep-deprived cohort.

In all three studies, I used neuroimaging techniques to identify diffusivity differences in the LC and the LC ascending noradrenergic bundle in publicly available datasets. Study 1 utilized the Berlin Aging Study-II (N = 301) and the Leipzig Study for Mind-Body-Emotion Interactions (N = 220) in attempt to replicate previous LC fractional anisotropy findings by Langley et al. (2020) in relation to the LC ascending noradrenergic bundle. Results show a successful replication of LC fractional anisotropy, with atypical age-related findings of younger adults displaying less fractional anisotropy relative to older adults. However, the LC ascending noradrenergic bundle displayed expected age-related changes in fractional anisotropy, with younger adults observed to have more fractional anisotropy than older adults. Additionally, our control region, the frontopontine tract that runs anterior to the LC, was observed to have no significant difference in fractional anisotropy between age groups. These findings suggest that diffusivity in the LC versus the LC ascending noradrenergic bundle may be susceptible to opposing microstructural changes throughout aging.

In study 2, I expanded the aims of our first study and utilized the Nathan Kline Institute (NKI) / Rockland dataset (N=369), to identify how diffusivity in the LC and the LC ascending noradrenergic bundle change throughout the lifespan, from early childhood (6 years old) to late life (75 years old). Our results replicate aim 1 age-related differences of fractional anisotropy of the LC and expand support for the hypothesis that diffusivity in the LC may be susceptible to opposing microstructural changes in relation to the LC ascending noradrenergic bundle. However, we were not able to replicate differences in fractional anisotropy in the LC ascending noradrenergic bundle between specific young and older adults age-groups. Yet, striking differences in axial, radial, and mean diffusivity that demonstrate age-related changes across the lifespan may be uniquely different in the LC compared to the LC ascending noradrenergic bundle.

Approximately 169 participants had two additional follow up visits for which I observed consistent diffusivity trends in each region of interest. Finally, I was not able to replicate ray-auditory verbal learning test (RAVLT) results identified previously with LC diffusivity but did identify associations with the LC noradrenergic ascending bundle. Together, these findings further demonstrate that LC diffusivity is unique in relation to the LC ascending noradrenergic bundle and frontopontine tracts. Finally, I discuss that using fixed age-ranges to study diffusivity changes may not be as informative as utilizing lifespan data.

In study 3, I explore what may drive the underlying mechanisms for the observed age-related differences in diffusivity between the LC and its ascending noradrenergic bundle. For this, I turned to the Stockholm Sleepy Brain dataset (N = 49), where young and older adults underwent a single night of sleep deprivation before their MRI scan. The results suggested that age-related differences in fractional anisotropy may be associated with sleep deprivation. Specifically, I was only able to replicate the increased fractional anisotropy in the LC of older adults when comparing sleep deprived participants. We further observed that sleep deprived young adults, relative to sleep rested young adults, had lower fractional anisotropy in the LC. But, sleep-deprived older adults did not show any difference in fractional anisotropy, compared with sleep rested older adults. This suggests that the LC of younger adults may be more susceptible to sleep deprivation effects than older adults. These findings, although limited in scope, may provide a framework for a future sleep-deprivation diffusion imaging study that seeks to replicate our results.

Together, the findings from these three studies provide evidence to suggest that the microstructural changes in the locus coeruleus are unique and do not adhere to typical age-related declines in diffusivity as observed with the LC ascending noradrenergic bundle and frontopontine tract. Additionally, the sleep-related observations support further cause for replication in larger cohorts, as they have broad implications to how sleep may impact the health of the locus coeruleus. Given the role of the LC in arousal and attention, as well as diseases like Alzheimer’s, the results in these studies provide a framework for further exploration of non-invasive neuroimaging biomarkers that may provide clinical benefits to the aging population.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

How age-related changes in the locus coeruleus affect selective attention and memory

PDF

Structure and function of the locus coeruleus across the lifespan

PDF

Noradrenergic mechanisms of arousal-enhanced memory selectivity

PDF

Value-based decision-making in complex choice: brain regions involved and implications of age

PDF

The role of the locus coeruleus in Alzheimer’s disease and cerebrovascular function: insights from neuroimaging, neuropsychology, and biofluid markers

PDF

Dedifferentiation of emotion regulation strategies in the aging brain: an MVPA investigation

PDF

Aging and emotion regulation in the judgment of facial emotion

PDF

Optimism when imagining the future in old age influences the positivity effect in memory

PDF

Neuroimaging markers of risk & resilience to brain aging and dementia

PDF

Crowding in peripheral vision

PDF

Mechanisms of stress effects on learning and decision making in younger and older adults

PDF

The evolution of decision-making quality over the life cycle: evidence from behavioral and neuroeconomic experiments with different age groups

PDF

Morphological and microstructural models of typical development in the hippocampus and white matter

PDF

Vascular contributions to brain aging along the Alzheimer's disease continuum

PDF

The role of age, mood, and time- perspective on the interpretation of emotionally ambiguous information

PDF

Anemia and white matter: a diffusion MRI analysis of the human brain

PDF

How emotional arousal influences memory and learning in younger and older adults

PDF

Emotion, attention and cognitive aging: the effects of emotional arousal on subsequent visual processing

PDF

The role of blood pressure variability in cognitive decline, cerebrovascular disease and Alzheimer’s disease

PDF

Unveiling the white matter microstructure in 22q11.2 deletion syndrome with diffusion magnetic resonance imaging

Asset Metadata

Creator Porat, Shai (author)

Core Title Age differences in diffusivity in the locus coeruleus and its ascending noradrenergic tract

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Neuroscience

Degree Conferral Date 2022-08

Publication Date 07/18/2022

Defense Date 05/03/2022

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag aging,diffusion,locus coeruleus,MRI,OAI-PMH Harvest

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Irimia, Andrei (committee chair), Braskie, Meredith (committee member), Kanowski, Scott (committee member), Mather, Mara (committee member)

Creator Email porat@usc.edu,shai73@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC111372148

Unique identifier UC111372148

Legacy Identifier etd-PoratShai-10848

Document Type Dissertation

Format application/pdf (imt)

Rights Porat, Shai

Type texts

Source 20220718-usctheses-batch-954 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu