Longitudinal White and Gray Matter Response to Precision Medicine-Guided Intervention for Alzheimer’s Disease

Abstract

Background:

Alzheimer’s disease (AD) is a debilitating condition that is widely known to adversely affect gray matter (GM) and white matter (WM) tracts within the brain. Recently, precision medicine has shown promise in alleviating the clinical and gross morphological trajectories of patients with AD. However, regional morphological changes have not yet been adequately characterized.

Objective:

Investigate regional morphological responses to a precision medicine-guided intervention with regards to white and gray matter in AD and mild cognitive impairment (MCI).

Methods:

Clinical and neuroimaging data were compiled over a 9-month period from 25 individuals who were diagnosed with AD or MCI receiving individualized treatment plans. Structural T1-weighted MRI scans underwent segmentation and volumetric quantifications via Neuroreader. Longitudinal changes were calculated via annualized percent change of WM or GM ratios.

Results:

Montreal Cognitive Assessment scores (p < 0.001) and various domains of the Computerized Neurocognitive Screening Vital Signs significantly improved from baseline to 9-month follow-up. There was regional variability in WM and GM atrophy or hypertrophy, but none of these observed changes were statistically significant after correction for multiple comparisons.

INTRODUCTION

Alzheimer’s disease (AD) is a debilitating condition that represents a spectrum of neurocognitive statuses ranging from mild cognitive impairment (MCI) to end-stage dementia [1]. Genetic factors that contribute to the development of AD include the presence of apolipoprotein E genes and methylenetetrahydrofolate reductase gene polymorphisms [2, 3]. There are also environmental factors in the development of AD, including cerebrovascular disease, traumatic brain injury, and cognitive activity [4]. Prevalent comorbidities such as diabetes, cardiovascular disease, and depression have also been associated with AD progression, potentially due to neuroinflammatory processes [5]. Novel efforts at curbing the progression of AD primarily revolve around disease-modifying therapies targeting well-known AD biomarkers such as amyloid-β and tau [6, 7]. Another well-explored interventional method to stem AD progression is the implementation of various lifestyle changes, including, but not limited to, diet, exercise, and nutritional supplementation [8 –10].

Recently, precision medicine has surfaced as a promising approach to preventing progression of AD, consisting of a multidisciplinary approach focusing on various lifestyle factors of individuals in addition to addressing their comorbidities [4]. A pilot study implementing this individualized approach demonstrated cognitive improvements in neuropsychological tests as well as decreased gross rates of brain atrophy [11]. However, detailed morphological trends, including those at the regional and subcortical levels, have not yet been investigated with regards to outcomes of precision medicine in AD. AD is well-described with regards to gray matter (GM) changes, but white matter (WM) density and integrity are measures of interest as demyelination may manifest in earlier stages of AD [12]. Well-known regions of WM damage in AD are localized to frontal, temporal, and parietal lobes, areas that functionally correspond to the clinical syndromes of AD [13, 14]. Though the mechanisms of these morphological changes are not currently well-understood, previous studies hypothesize that WM damage may occur in a multitude of pathways, including secondary to Wallerian degeneration, or independently of gray matter changes via retrogenesis [15 –17]. Thus, investigating WM changes may provide a deeper perspective into the therapeutic processes of precision medicine-guided interventions in AD.

This study aims to elucidate a more detailed insight into the morphological trends in the brain in response to precision medicine-based intervention, particularly with regards to WM changes and their localization. We hypothesize that paralleling the gross morphological trends that was observed in the previous study, WM changes in the cohort will be more reflective of normal aging patterns than traditional AD trajectories.

METHODS

Participants and evaluation

25 individuals who were diagnosed with AD or MCI from three clinical sites, Walnut Creek, California; San Rafael, California; and Ashland, Oregon, were selected for the 4 Winds Precision Medicine approach to Alzheimer’s Disease trial [11]. Each participant underwent treatment on an individualized basis consisting of diet, exercise, sleep hygiene, stress management, and supplements. If applicable, participants underwent treatment of comorbid infectious or toxic etiologies associated with cognitive decline or systemic inflammation, as well as correction of hormonal and endocrine deficiencies. Though detailed treatment protocol for each condition is more thoroughly explained in the pilot study, interventions of note included continuous positive airway pressure for participants with sleep apnea or upper airway resistance syndrome, valacyclovir for 2-6 months for those with evidence of Herpes simplex infection, and herbal remedies for those with active Epstein-Barr Virus or tick-borne infections [11].

Neuropsychological screening data was characterized with a combination of tools including Montreal Cognitive Assessment (MoCA) and Alzheimer’s Questionnaire (AQ-21), as well as Computerized Neurocognitive Screening Vital Signs (CNS-VS), a neurocognitive test battery represented by standardized scores with a mean of 100 and a standard deviation of 15. The CNS-VS were categorized into domains consisting of memory, verbal memory, visual memory, psychomotor speed, reaction time, processing speed, executive functioning, simple attention, and motor speed. Participants’ cognitive status was categorized based on AQ-21 scores, with previous validation of this tool demonstrating that those with scores 5 to 14 typically had MCI, whereas scores greater than 14 were associated with dementia [18]. With this stratification, 22 participants were diagnosed as having MCI and 3 were diagnosed as having dementia. Detailed protocols of participant selection and characterization, including inclusion and exclusion criteria, are described further in the prior publication [11]. In short, from a medical standpoint, exclusion criteria consisted of major health conditions such as seizures, cardiovascular disease, cancer, major psychiatric diagnosis, history of deep vein thrombosis, stroke, and diagnosis of a neurodegenerative disease other than AD (e.g., frontotemporal dementia). Concerns of practice effects were addressed by administering different MoCA tests at baseline, 3-month, 6-month, and 9-month evaluations and implementing the CNS-VS, which has been experimentally proven to minimize such effects [19].

Imaging

Each participant underwent magnetic resonance imaging (MRI) of the brain at baseline and after 9 months, upon completion of the treatment protocol. All scans were retrieved via 3-Tesla MRI scanners with either magnetization-prepared rapid acquisition gradient echo or spoiled gradient-recalled echo sequences. Each scan underwent segmentation and volumetric quantifications via Neuroreader, v2.7.0 [20]. In summary, the pipeline utilizes multi-atlas segmentation through non-linear registration and the resulting segmentations are run through multiple layers of corrections. Hippocampal segmentation was similarly achieved with a multi-atlas approach and probability maps were created with multiple templates to account for anatomic variabilities. Furthermore, in the novel version of the pipeline, lobar regions were further segmented into GM and WM volumes.

Statistical analyses

Volumetric change between baseline and follow-up were quantified via annualized percent change. Each volumetric measurement was corrected by the individual’s mean total intracranial volume and normalized into a percentage ratio. Paired t-tests were performed with raw annualized change rates of brain matter volumes to determine significance between baseline and 9-month follow-up at a significance level of 0.05. Wilcoxon sign-rank tests were utilized to compare overall MoCA scores and subscores between baseline and 9-month follow-up. All statistical analyses were performed with SPSS, Version 26. Corrections for multiple comparisons were performed using the Benjamini-Hochberg procedure at a false discovery rate of 0.05 [21].

RESULTS

Neuropsychological testing

A summary of demographics and neuropsychological testing results is listed in Table 1. MoCA scores (p < 0.001) and memory (p = 0.023), verbal memory (p = 0.006), psychomotor speed (p = 0.001), reaction time (p = 0.009), executive functioning (p < 0.001), simple attention (p = 0.023), and motor speed (p < 0.001) domains of the CNS-VS significantly improved from baseline to 9-month follow-up. Visual memory (p = 0.939) and processing speed (p = 0.166) domains of the CNS-VS did not significantly differ between the two time points.

Table 1
Demographic and neuropsychological information

Demographic (n = 25)

Sex

Male 12

Female 13

Age 50–76

Ethnicity

Caucasian 24

Asian American 1

Education

High school 1

Some college 2

College graduate 11

Post-graduate 11

Category Baseline 9mo Difference Sig

MoCA 24.88 27.56 2.68 0.0009

CNS-VS

Memory 87.08 96.16 9.08 0.023

Verbal memory 85.16 99.44 14.28 0.006

Visual memory 93.96 94.04 0.08 0.939

Psychomotor speed 98.44 106.68 8.24 0.001

Reaction time 93.56 101.04 7.48 0.009

Processing speed 101.28 104.96 3.68 0.166

Executive function 97.48 109.68 12.2 0.0001

Simple attention 90.96 101.48 10.52 0.024

Motor speed 97.6 106 8.4 0.0008

Demographic (n = 25)
Sex
Male	12
Female	13
Age	50–76
Ethnicity
Caucasian	24
Asian American	1
Education
High school	1
Some college	2
College graduate	11
Post-graduate	11
Category	Baseline	9mo	Difference	Sig
MoCA	24.88	27.56	2.68	0.0009
CNS-VS
Memory	87.08	96.16	9.08	0.023
Verbal memory	85.16	99.44	14.28	0.006
Visual memory	93.96	94.04	0.08	0.939
Psychomotor speed	98.44	106.68	8.24	0.001
Reaction time	93.56	101.04	7.48	0.009
Processing speed	101.28	104.96	3.68	0.166
Executive function	97.48	109.68	12.2	0.0001
Simple attention	90.96	101.48	10.52	0.024
Motor speed	97.6	106	8.4	0.0008

MoCA, Montreal Cognitive Assessment (max score = 30); CNS-VS, Computerized Neurocognitive Assessment Vital Signs (standardized scores with mean = 100 and standard deviation = 15); 9mo, 9 months; Sig, significance.

White matter volume changes

A comprehensive list of WM volume changes is listed in Table 2. Cerebral WM volume decreased at a rate of 1.467% /year. Reductions were most pronounced in the frontal lobe regions (left: -1.393% /year; right: -4.995% /year), while increased WM volumes were most observed in the occipital (left: 0.304% /year; right: 2.104% /year) and parietal lobes (left: 1.952% /year, right: 0.611% /year). None of the regional WM volume changes were statistically significant after correction for multiple comparisons.

Table 2

Cross-sectional and longitudinal data of White Matter Volume/mTIV ratios

White Matter Volume/mTIV Ratio % (n = 25)
Region	Baseline	9mo	Raw Change	Annualized % Change (% /year)
Frontal lobe, left	5.148	5.097	–0.051	–1.393
Frontal lobe, right	5.197	5.012	–0.185	–4.995
Parietal lobe, left	2.733	2.771	0.038	1.952
Parietal lobe, right	2.710	2.722	0.012	0.611
Temporal lobe, left	2.512	2.472	–0.040	–2.249
Temporal lobe, right	2.481	2.472	–0.008	–0.47
Occipital lobe, left	1.251	1.253	0.003	0.304
Occipital lobe, right	1.246	1.264	0.019	2.104
Cerebrum, left	11.643	11.593	–0.051	–0.61
Cerebrum, right	11.634	11.471	–0.163	–1.964
Cerebellum, left	1.073	1.082	0.009	1.153
Cerebellum, right	1.086	1.068	–0.018	–2.356
Overall white matter	27.621	27.332	–0.289	–1.467

mTIV, mean total intracranial volume; 9mo, 9 months.

3.3 Gray matter volume changes

A comprehensive list of GM volume changes is listed in Table 3. Cerebral GM volume decreased at a rate of 0.064% /year. Reductions were most pronounced in the right occipital (-5.92% /year) and parietal (left: -3.138% /year, right: -3.631% /year) regions, while increased GM volumes were most prominently observed in the right frontal (4.548% /year) and left temporal (1.972% /year) regions, though raw changes were less pronounced. None of the regional GM volume changes were statistically significant after correction for multiple comparisons.

Table 3

Cross-sectional and longitudinal data of Gray Matter Volume/mTIV ratios

Gray Matter Volume/mTIV Ratio % (n = 25)
Region	Baseline	9mo	Raw Change	Annualized % Change (% /year)
Frontal lobe, left	5.425	5.429	0.003	0.083
Frontal lobe, right	5.424	5.600	0.176	4.548
Parietal lobe, left	2.591	2.533	–0.058	–3.138
Parietal lobe, right	2.617	2.550	–0.068	–3.631
Temporal lobe, left	3.356	3.403	0.047	1.972
Temporal lobe, right	3.501	3.457	–0.044	–1.756
Occipital lobe, left	1.448	1.423	–0.025	–2.419
Occipital lobe, right	1.345	1.288	–0.057	–5.92
Cerebrum, left	12.819	12.787	–0.033	–0.356
Cerebrum, right	12.887	12.895	0.008	0.082
Cerebellum, left	2.667	2.671	0.004	0.235
Cerebellum, right	2.725	2.723	–0.001	–0.073
Overall gray matter	32.975	32.960	–0.015	–0.064

mTIV, mean total intracranial volume; 9mo, 9 months.

DISCUSSION

Previous studies show structural WM reductions of 0.5% -1% /year in normal aging, in a diffusively homogeneous pattern [22]. Precise annualized rates of WM reductions in AD are not well-elucidated in previous literature, but most significantly affected regions include bilateral medial temporal lobes [23]. In this study, many regions, along with overall WM volume, endured a change closer to those in normal aging trajectories than in AD, supporting our initial hypothesis. Furthermore, though not statistically significant, there were measurable increases in WM volumes in certain areas, especially the occipital lobe, which may be attributed to the tendency that AD preferentially affects the WM of frontal, parietal, and temporal regions [13, 14]. These findings are also accompanied by improved neuropsychological screening testing results in multiple domains of the CNS-VS, which have been corroborated in the pilot study [11]. Additionally, these differential increases may be attributed to compensatory reallocation of cognitive resources to different brain regions, as suggested by previous fMRI and connectome studies [24, 25].

GM atrophy rates generally range from 0.5-1% /year in normal aging, and 1-4% /year in AD, in areas congruous to WM atrophy [26, 27]. Regional improvements in GM localized around areas of AD-specific regions, particularly the frontal and temporal lobes, though there was noticeable variability between hemispheres. Although the mechanisms behind the observed regional variances in GM atrophy are unclear at this point, given that occipital lobe atrophy is not known to be associated with AD pathology, and both GM and WM volumes within AD-specific regions (frontal, parietal, temporal lobes) did not significantly decrease, a precision-medicine guided approach may align morphological changes to correspond more closely with normal aging than AD [28]. Further neurobiological studies should be pursued to investigate in detail the mechanisms behind the regional effects of a combined therapeutic approach.

Interestingly, the WM and GM changes exhibited some apparent decoupling, particularly in the occipital, frontal, and temporal lobes. Regions that had greater gray matter reductions tended to exhibit greater white matter increases, and vice versa. In individuals with AD, cortical damage and WM degeneration are spatially correlated [27]. This study suggests that precision medicine-based intervention may contribute in breaking down this relationship, as such associations are inconclusive in normal aging or even MCI [29, 30]. Thus, a precision medicine-based approach may curtail the pathophysiological processes driving the congruent reductions of WM and GM in AD, such as rarefaction of myelin, axonal gliosis, and Wallerian degeneration [29]. However, the precise mechanism of the apparent quantitative inverse relationship between WM and GM changes remains uncertain, and future studies should explore this potential phenomenon.

There exist limitations within our study, many of which are shared with the pilot study [11]. Though the study benefits from being longitudinal in nature, the small sample size of 25 individuals rendered statistical analyses to be less powerful than desired, and initially statistically significant results were marred by correction of multiple comparisons. Additionally, the cognitive screening tools and their respective boundaries utilized in this study have limited validity and reliability in diagnosing disease presence and severity compared to standardized, multi-hour, comprehensive neuropsychological evaluations. Particularly, though statistically significant, the observed neuropsychological testing score improvements may be partly attributed to ceiling effects, natural variability, and regression to the mean. Therefore, these observed raw score improvements of the MoCA scores and the CNS-VS may not necessarily translate to clinically significant outcomes, and should be further explored by employing a more comprehensive evaluation of objective and subjective findings. Another limitation was that imaging was only obtained at baseline and the 9-month follow-up period, which combined with its small timeframe relative to the chronic course of AD, limits generalization of the temporality of the study’s results. This limitation may have also played a factor in the high variability of regional differences observed in atrophy or hypertrophy rates. Furthermore, the nature of the calculation of annualized differences, given the relatively small volume ratios of some of the regions, may have over or underestimated the precise magnitude of such changes. Finally, the lack of a control group limits assessment of the relative efficacy of this intervention compared to other established treatment modalities.

Despite limitations, when considering the imaging results with observed functional and cognitive improvements, precision medicine-based interventions carry potential in addressing the neurodegenerative processes of AD. It is already known that generalized lifestyle changes, such as physical activity, mental activity, and adherence to a Mediterranean-based diet reduces brain atrophy rates in various AD-susceptible regions [31 –33]. Such lifestyle changes, as well as addressing hormonal deficiencies and infections, have also been shown to reverse age-related brain volume losses and decrease the risk of AD progression [34, 35]. Moreover, comorbidities such as type II diabetes, metabolic syndrome, and obstructive sleep apnea have been shown to either contribute to neurodegenerative processes of AD or share the same pathophysiological mechanisms as AD, demonstrating a potential for this precision medicine-guided intervention to actively treat AD pathology [5 , 37]. Thus, combining an individualized medical plan with generalized lifestyle recommendations may be an important step to improve both clinical and morphological outcomes of AD. Future innovations of therapy should be targeted in generating consistent levels of neurogenesis that correlate with significant increases in volumes and cognitive status.

AD is one of the most chronically debilitating diseases that commonly affect individuals with advancing age and is a condition that does not have a reliable form of treatment. However, pilot studies have established a precision medicine-based intervention as a promising approach to curtailing the pathophysiological processes in AD, both clinically and morphologically. This study validates that with the precision medicine intervention used in this pilot trial, there was no significant brain atrophy with regards to WM or GM in this cohort, though there are subtle regional variabilities that additional studies should further investigate across a longer time period. These findings support the potential efficacy of an individualized approach to the treatment of AD and may align individuals with AD to a clinical and morphological trajectory more fitting of normal aging.

Footnotes

ACKNOWLEDGMENTS

Outside contributions remain unchanged from the pilot study. We acknowledge the Four Winds Foundation, Diana Merriam, and Gayle Brown for their support of this study. We are grateful to David Perlmutter for introductions and CNS Vital Signs, Posit Science, HeartMath, Great Plains Laboratory, Cyrex Laboratory, Armin Labs, Bio-Botanical Research, Genova Diagnostics, Doctors Data, IntellxxDNA, RadNet, and Norcal Imaging for providing diagnostics for the study. We also thank Brainreader ApS for providing the Neuroreader imaging software for analyses. For clinical support, we acknowledge Tereasa Simonson MD, Michael Atkinson, Derek Barber, Mac Dodds, James Gaydos DO, Doug Jaser, Lynn Kilips, Karen Preskenis, Renèe Riley-Adams, Venessa Rodriguez, Kia Sanford, and Sheila Wagner ND. We are grateful to Phyllis and James Easton for research support leading to this trial.

FUNDING

The project was funded from the Four Winds Foundation, a non-profit entity, with no additional sources of funds. The funding entity had no role in the study design, data collection, data analysis, data interpretation, or writing of the report. None of the authors received compensation for the study except from the Foundation, to the 3 performing physicians for seeing the patients, and the contract research organization (Quesgen).

CONFLICT OF INTEREST

Dr. Cyrus Raji is an Editorial Board Member of this journal but was not involved in the peer-review process nor had access to any information regarding its peer-review. He received consulting fees from Brainreader ApS, Apollo Health, Neurevolution LLC, Voxelwise LLC, and the PNI Foundation. Dr. Kat Toups accepted lecture fees from BioBotanical Research ($1,000) in 2021. Dr. Mouna Attarha is an employee and senior research scientist at Posit Science, the developer of BrainHQ. and is a shareholder of Posit Science stock. Dr. Dale Bredesen received consulting fees from Life Seasons ($10,0000 per month) and Apollo Health ($10,000 per month), and owns equity in Apollo Health (no current value). Dr. Sharon Hausman-Cohen is the Medical Director of IntellxxDNA which was the genomics platform used in the study. We did not sponsor the study but did provide the genomics at a discount. Michael Jarrett owns equity of the company that provided data management services and support for study operations for this study. He owns other equities in his personal portfolio, but none that pertain to the execution of this study. All other authors have no conflict of interest to report.

DATA AVAILABILITY

The data supporting the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

References

McKhann

, Knopman

, Chertkow

, Hyman

, Jack

Jr. , Kawas

, Klunk

, Koroshetz

, Manly

, Mayeux

, Mohs

, Morris

, Rossor

, Scheltens

, Carrillo

, Thies

, Weintraub

, Phelps

(2011) The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 263–269.

Ballard

, Gauthier

, Corbett

, Brayne

, Aarsland

, Jones

(2011) Alzheimer’s disease. Lancet 377, 1019–1031.

Román

(2015) MTHFR gene mutations: A potential marker oflate-onset Alzheimer’s disease? J Alzheimers Dis 47, 323–327.

Behl

, Kaur

, Sehgal

, Singh

, Albarrati

, Albratty

, Najmi

, Meraya

, Bungau

(2022) The road to precision medicine: Eliminating the “One Size Fits All” approach in Alzheimer’s disease. Biomed Pharmacother 153, 113337.

Santiago

, Potashkin

(2021) The impact of disease comorbidities in Alzheimer’s disease. Front Aging Neurosci 13, 631770.

Long

, Holtzman

(2019) Alzheimer disease: An update on pathobiology and treatment strategies. Cell 179, 312–339.

Cummings

, Lee

, Ritter

, Sabbagh

, Zhong

(2020) Alzheimer’s disease drug development pipeline: 2020. Alzheimers Dement (N Y) 6, e12050.

Lövdén

, Xu

, Wang

(2013) Lifestyle change and theprevention of cognitive decline and dementia: What is the evidence? Curr Opin Psychiatry 26, 239–243.

Barnard

, Bush

, Ceccarelli

, Cooper

, de Jager

, Erickson

, Fraser

, Kesler

, Levin

, Lucey

, Morris

, Squitti

(2014) Dietary and lifestyle guidelines for the prevention of Alzheimer’s disease. Neurobiol Aging 35 Suppl 2, S74–78.

10.

Shea

, Remington

(2015) Nutritional supplementation for Alzheimer’s disease? Curr Opin Psychiatry 28, 141–147.

11.

Toups

, Hathaway

, Gordon

, Chung

, Raji

, Boyd

, Hill

, Hausman-Cohen

, Attarha

, Chwa

, Jarrett

, Bredesen

(2022) Precision medicine approach to Alzheimer’s disease: Successful pilot project. J Alzheimers Dis 88, 1411–1421.

12.

Sachdev

, Zhuang

, Braidy

, Wen

(2013) Is Alzheimer’s a disease of the white matter? Curr Opin Psychiatry 26, 244–251.

13.

Bozzali

, Falini

, Franceschi

, Cercignani

, Zuffi

, Scotti

, Comi

, Filippi

(2002) White matter damage in Alzheimer’s disease assessed using diffusion tensor magnetic resonance imaging. J Neurol Neurosurg Psychiatry 72, 742–746.

14.

Duan

, Wang

, Xu

, Lin

, Chen

, Kang

, Yao

(2006) White matter damage of patients with Alzheimer’s disease correlated with the decreased cognitive function. Surg Radiol Anat 28, 150–156.

15.

Amlien

, Fjell

(2014) Diffusion tensor imaging of white matter degeneration in Alzheimer’s disease and mild cognitive impairment. Neuroscience 276, 206–215.

16.

Bartzokis

, Lu

, Mintz

(2007) Human brain myelination and amyloid beta deposition in Alzheimer’s disease. Alzheimers Dement 3, 122–125.

17.

Mayo

, Mazerolle

, Ritchie

, Fisk

, Gawryluk

(2017) Longitudinal changes in microstructural white matter metrics in Alzheimer’s disease. Neuroimage Clin 13, 330–338.

18.

Malek-Ahmadi

, Sabbagh

(2015) Development and Validation of the Alzheimer’s Questionnaire (AQ). J Nat Sci 1, e104.

19.

Field

, Edwards

, Dean

(2018) Influence of HeartMath Training Programme on physiological and psychological variables. Glob J Health Sci 10 doi:10.5539/gjhs.v10n2p126.

20.

Ahdidan

, Raji

, DeYoe

, Mathis

, Noe

, Rimestad

, Kjeldsen

, Mosegaard

, Becker

, Lopez

(2016) Quantitative neuroimaging software for clinical assessment of hippocampal volumes on MR imaging. J Alzheimers Dis 49, 723–732.

21.

Benjamini

, Hochberg

(1995) Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Stat Soc Series B Methodol 57, 289–300.

22.

Schilling

, Archer

, Yeh

, Rheault

, Cai

, Hansen

, Yang

, Ramdass

, Shafer

, Resnick

, Pechman

, Gifford

, Hohman

, Jefferson

, Anderson

, Kang

, Landman

(2022) Aging and white matter microstructure and macrostructure: A longitudinal multi-site diffusion MRI study of 1218 participants. Brain Struct Funct 227, 2111–2125.

23.

, Pan

, Huang

, Shang

(2012) A meta-analysis of voxel-based morphometry studies of white matter volume alterations in Alzheimer’s disease. Neurosci Biobehav Rev 36, 757–763.

24.

Wang

, Liang

, Wang

, Tian

, Zhang

, Li

, Jiang

(2007) Altered functional connectivity in early Alzheimer’s disease: A resting-state fMRI study. Hum Brain Mapp 28, 967–978.

25.

Filippi

, Basaia

, Canu

, Imperiale

, Magnani

, Falautano

, Comi

, Falini

, Agosta

(2020) Changes in functional and structural brain connectome along the Alzheimer’s disease continuum. Mol Psychiatry 25, 230–239.

26.

Anderson

, Schott

, Bartlett

, Leung

, Miller

, Fox

(2012) Gray matter atrophy rate as a marker of disease progression in AD. Neurobiol Aging 33, 1194–1202.

27.

Agosta

, Pievani

, Sala

, Geroldi

, Galluzzi

, Frisoni

, Filippi

(2011) White matter damage in Alzheimer disease and its relationship to gray matter atrophy. Radiology 258, 853–863.

28.

Middelkoop

, van der Flier

, Burton

, Lloyd

, Paling

, Barber

, Ballard

, McKeith

, O’Brien

(2001) Dementia with Lewy bodies and AD are not associated with occipital lobe atrophy on MRI. Neurology 57, 2117–2120.

29.

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. Neuropsychiatr Dis Treat 4, 737–742.

30.

Giorgio

, Santelli

, Tomassini

, Bosnell

, Smith

, De Stefano

, Johansen-Berg

(2010) Age-related changes in grey and white matter structure throughout adulthood. Neuroimage 51, 943–951.

31.

Smith

, Nielson

, Woodard

, Seidenberg

, Durgerian

, Hazlett

, Figueroa

, Kandah

, Kay

, Matthews

, Rao

(2014) Physical activity reduces hippocampal atrophy in elders at genetic risk for Alzheimer’s disease. Front Aging Neurosci 6, 61.

32.

Valenzuela

, Sachdev

, Wen

, Chen

, Brodaty

(2008) Lifespan mental activity predicts diminished rate of hippocampal atrophy. PLoS One 3, e2598.

33.

Ballarini

, Melo van Lent

, Brunner

, Schröder

, Wolfsgruber

, Altenstein

, Brosseron

, Buerger

, Dechent

, Dobisch

, Duzel

, Ertl-Wagner

, Fliessbach

, Freiesleben

, Frommann

, Glanz

, Hauser

, Haynes

, Heneka

, Janowitz

, Kilimann

, Laske

, Maier

, Metzger

, Munk

, Perneczky

, Peters

, Priller

, Ramirez

, Rauchmann

, Roy

, Scheffler

, Schneider

, Spottke

, Spruth

, Teipel

, Vukovich

, Wiltfang

, Jessen

, Wagner

(2021) Mediterranean diet, Alzheimer disease biomarkers and brain atrophy in old age. Neurology 96, e2920–2932.

34.

Meeusen

(2014) Exercise, nutrition and the brain. Sports Med 44 Suppl 1, S47–56.

35.

Sandison

, Callan

NGL

, Rao

, Phipps

, Bradley

(2023) Observed improvement in cognition during a personalized lifestyle intervention in people with cognitive decline. J Alzheimers Dis 94, 993–1004.

36.

Andrade

, Bubu

, Varga

, Osorio

(2018) The relationship between obstructive sleep apnea and Alzheimer’s disease. J Alzheimers Dis 64, S255–S270.

37.

Čater

, Hölter

(2022) A pathophysiological intersectionof diabetes and Alzheimer’s disease. Int J Mol Sci 23, 11562.