A multi-modal medical management and lifestyle intervention increase cerebral blood flow and lowers diabetic risk in persons with early Alzheimer's disease: Mid-trial results from the PREVENTION trial

Abstract

Background

Medical and lifestyle management are crucial for Alzheimer's disease (AD). Cerebral blood flow (CBF), vital for brain health, and influenced by modifiable risk factors, is reduced in AD and may become uncoupled from metabolism due to neurovascular dysfunction in later stages.

Objective

This mid-trial analysis tested the hypothesis that a coached, multi-modal intervention (PREVENTION) improved ASL-MRI-measured CBF and diabetic risk (QUICKI) in patients with early AD.

Methods

The control arm received recommendations and medical management for one year; the active arm additionally received coaching, exercise training, and supplementation. We hypothesized that those in (1) the active arm and (2) with higher intervention adherence would have improved post-trial QUICKI and CBF, particularly in regions relevant to exercise, cardiovascular, diabetic, and AD risk. Post-trial CBF was analyzed using a linear model including arm, baseline CBF, adherence, age, education, and depressive symptoms. Change in QUICKI was analyzed using mixed effects general linear models, including arm, adherence, time, and interactions between time and treatment group and time and adherence, controlling for age.

Results

The active arm (n = 18) showed greater post-trial CBF in regions related to exercise, cardiovascular, diabetic, and AD risk, compared to control (n = 20), but did not differ in global CBF, QUICKI, or adherence. Higher adherence scores were associated with greater regional post-trial CBF and improvement in QUICKI, but not global CBF.

Conclusions

In this small sample, we found evidence that a multi-modal intervention focused on medical management, exercise, and a carbohydrate-restricted diet improved diabetic risk and CBF in patients with AD.

Keywords

Alzheimer's disease amyloid-β cerebral blood flow cognitive impairment diet lifestyle magnetic resonance imaging non-pharmacological interventions physical activity risk factors

Introduction

Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by cognitive decline and memory loss, leading to dementia. It represents the most common form of dementia, affecting millions globally.¹ This growing public health burden underscores the urgent need for effective interventions to prevent, slow progression, or manage symptoms.

Recent advancements have brought lecanemab and donanemab, which are approved by the FDA, considered disease-modifying immunotherapies for mild cognitive impairment (MCI) and early AD,^2,3 and represent a new class of drugs targeting the buildup of amyloid beta plaques in the brain. While these medications demonstrate promise in slowing cognitive decline, they are expensive and not yet widely covered by Medicare.^4,5 Importantly, amyloid-related imaging abnormalities (ARIA) are thought to cause side effects like headache, confusion, nausea, dizziness, and seizure. Risk of developing ARIA can exclude up to 90% of patients seeking treatment with new anti-amyloid monoclonal antibodies, especially those with existing vascular disease.⁶ Therefore, it is plausible that improving cerebrovascular health could mitigate these side effects. Additionally, despite these advancements, common treatment options for AD continue to focus on temporary symptom management rather than halting the disease process.⁵

The challenges associated with lecanemab and donanemab, including the high cost, limited coverage, large exclusion rates, potential medication side effects like ARIA, and the lack of alternative disease-modifying therapies, highlight the importance of exploring complementary approaches, such as multi-modal, health-centered interventions. The 2024 Lancet Commission update emphasizes growing evidence that addressing cardiovascular and diabetic risk factors, including hypertension, smoking, obesity, physical inactivity, cholesterol, and diabetes, plays a critical role in reducing dementia risk.¹ Several clinical trials that address modifiable risk factors through diet, exercise, and other interventions, such as the Finnish Geriatric Intervention Study to Prevent Cognitive Impairment (FINGER) trial, have shown positive results in preventing cognitive decline in high-risk individuals.⁷ The Coaching for Cognition in Alzheimer's (COCOA) trial found that a regularly coached, multi-modal lifestyle intervention emphasizing the Mediterranean-DASH Intervention for Neurodegenerative Delay (MIND) diet, physical activity recommendations, and providing cognitive training benefitted cognition in individuals with cognitive complaints.⁸ Lifestyle Intervention for Early Alzheimer's Disease (LIEAD) studied the effects of an intensive lifestyle intervention with plant-based, provided meals which excluded refined carbohydrates and sugars, along with supervised aerobic and strength training, stress management, and social support in patients with mild cognitive impairment (MCI) or early AD. They found improved cognition, functional outcomes, and plasma Aβ_42/40 ratio.⁹ Ongoing trials like the U.S. Study to Protect Brain Health Through Lifestyle Intervention to Reduce Risk (US POINTER) and World-Wide-FINGERS are investigating varied interventions addressing similar underlying concerns.^10,11

The Precision Recommendations for Environmental Variables, Exercise, Nutrition, Training Intervention to Optimize Neurocognition (PREVENTION) study is an ongoing randomized clinical trial (RCT)¹² aimed at targeting modifiable risk factors to reduce disease progression in patients with early-stage AD symptoms and confirmed amyloid neuropathology. The intervention included a carbohydrate-restricted MIND diet, multi-component physical exercise, cognitive training, stress reduction, medical management, and nutritional supplementation. The MIND diet is a hybrid of two well-researched diets—the Mediterranean diet and the DASH diet (Dietary Approaches to Stop Hypertension).¹³ While both arms received personalized multi-modal lifestyle recommendations and four medical visits, the active arm also received dietary counseling, group physical and cognitive exercise, health coaching, and nutritional supplements free of charge. All aspects of the intervention were personalized, informed by the patients’ medical history, assessments, clinical blood labs, physical activity levels, body mass index, and apolipoprotein gene (APOE) status.¹² Intervention components were emphasized based on personal patient data. The intervention arm also received coaching to refine personalized plans by incorporating each patient's ongoing functional abilities, support system, and motivation.

The goal of this mid-trial analysis is to explore the benefits of the PREVENTION Trial on exploratory outcomes absolute cerebral blood flow (CBF), and diabetic risk. No study has demonstrated the efficacy of a multi-modal, health-centered intervention focused on diet, exercise, supplementation, and medical management on CBF in patients with early AD and confirmed amyloidosis.

We selected CBF as an outcome for this exploratory study because arterial spin labeling (ASL) is increasingly recognized as a leading indicator of brain dysfunction in AD. It is a non-invasive marker of cerebrovascular health¹⁴ and one of the earliest pathological changes in AD. Reduced CBF occurs before significant amyloid-beta deposition or glucose metabolism impairments¹⁴ and precedes and potentially drives neurodegeneration.¹⁵ ASL's ability to capture these early alterations makes it a valuable biomarker for tracking AD progression in clinical research.

We hypothesized that participants in the active arm would exhibit higher post-trial CBF, particularly in brain regions previously linked to physical activity, insulin sensitivity, and cardiovascular risk, and known to be relevant to AD.^16–23 Aerobic physical activity has been associated with increased perfusion in the medial temporal lobe (MTL) and posterior cingulate gyrus.^18,21,22 Insulin resistance and reduced insulin sensitivity have been linked to widespread cortical perfusion deficits, with regional findings reported in the anterior cingulate, medial and middle frontal gyri, inferior parietal lobule, orbitofrontal gyrus, precentral gyrus, precuneus, posterior cingulate, middle occipital gyrus, superior frontal and parietal cortex^16,17 Cardiovascular risk has been associated with lowered perfusion in the anterior cingulate, medial and middle frontal gyri, precentral gyrus, middle temporal gyrus, and medial superior frontal gyrus, with broader effects observed across the temporal lobe.^17,23 Moreover, we hypothesized that participants in the active arm would demonstrate more stable or improved diabetic risk. We also hypothesized that the level of adherence, independent of treatment arm and confounders, would be associated with better outcomes across all measures.

Methods

Participants

Participants are drawn from the high-volume Pacific Brain Health Center (PBHC) memory care clinic, which has the benefit of assessing feasibility in a real-world clinical setting. PBHC is located in Santa Monica, California, with referrals from the surrounding area, including Los Angeles County. PBHC is part of Providence St Joseph Health system.¹² The study was approved by WIRB-Copernicus Group Institutional Review Board (Protocol # 20190583). Informed consent was obtained from each of the participants. This research was conducted in accordance with the Helsinki Declaration of 1975.

Eligibility criteria required participants to be ≥50 years old and at Functional Assessment Staging Tool for Dementia (FAST)²⁴ Stages 2–4 (subjective cognitive decline, MCI or early AD). Those at FAST Stage 4 were required to have a caregiver or legally appointed representative willing to assist with required procedures. Participants were required to be amyloid positive by PET, CSF, or blood-based testing, fluent in English, able to use a computer, and capable of telephonic or video communication. Participants also needed normal or corrected vision and hearing, and either answered “no” to all items of the Physical Activity Readiness Questionnaire²⁵ or obtained physician clearance to participate in a moderately intensive exercise program. Exclusion criteria included individuals with clinical presentation suggestive of non-AD neurodegenerative disorder (e.g., Lewy body dementia, frontotemporal dementia), those whose primary cause of cognitive impairment was attributed to cerebrovascular disease, those with immediate family members carrying known AD mutation in the PSEN or APP genes, and individuals with a Mini-Mental Status Exam below 19 or Clinical Dementia Rating Scale ≥2, as noted in their patient medical record.

Screening of individuals for the study for biomarker evidence of AD²⁶ was done by either positron emission tomography imaging with Florbetapir (18F),²⁷ cerebrospinal fluid amyloid,²⁸ or the PrecivityAD2 blood test.²⁹ All individuals enrolled in the study prior to our data freeze (December 31, 2023). Patients were included in the imaging analysis if they received ASL and T1-weighted structural magnetic resonance imaging (sMRI) at both baseline and 12-month follow-up post-trial. Two participants were excluded due to contraindications for receiving sMRI. Eighteen were excluded due to lack of follow-up sMRI data.

Participants or their care partners provided a detailed medical history along with demographic information, including sex, date of birth, handedness, ethnicity, and race. See Table 1 for participant demographics. Medical history was confirmed through manual review of patient medical records through an Electronic Health Record system.

Table 1.

Demographic and baseline characteristics.

	Control (n = 30)	Active (n = 31)	Total (n = 61)
Sex (n (% female))	12 (40)	16 (51.6)	28 (45.9)
Age (y)
Mean (SD)	74.0 (7.3)	72.8 (7.2)	73.4 (7.2)
Range (min-max)	57.0–89.2	59.3–87.0	57.0–89.2
Education (y)
Mean (SD)	17.0 (2.5)	16.9 (3.1)	16.9 (2.8)
Range (min-max)	12.0–20.0	12.0–23.0	12.0–23.0
Handedness (% right)	80.0	96.8	88.5
FAST Stage
Mean (SD)	3.2 (0.5)	3.2 (0.4)	3.2 (0.4)
Range (min-max)	2–4	3–4	2–4
Depressive symptoms
Mean (SD)	50.5 (8.2)	49.5 (6.3)	50.5 (7.3)
Range (min-max)	(41.0–69.5)	(41.0–64.3)	(41.0–60.0)
Ethnicity (n (% Hispanic or Latino))	0 (0.0)	2 (6.5)	2 (3.3)
Race (n (% Black or African American))	0 (0.0)	1 (3.2)	1 (1.6)
Race (n (% Asian or Native Hawaiian or Pacific Islander))	3 (10.0)	4 (12.9)	7 (11.5)
Race (n (% White))	27 (90.0)	25 (80.6)	52 (85.2)
Race (n (% Other))	0 (0.0)	1 (3.2)	1 (1.6)
APOE ε4 carrier (n (%))	22 (73.3)	22 (71.0)	44 (72.1)
QUICKI (n)	27	30	57
Mean (SD)	0.36 (0.03)	0.36 (0.04)	0.36 (0.04)
Range (min-max)	0.32–0.45	0.30–0.43	0.30–0.45
Baseline global CBF (n)	24	26	50
Mean (SD)	27.2 (6.7)	22.8 (5.4)	24.8 (7.1)
Range (min-max)	13.0–40.9	13.9–35.0	13.0–40.9

FAST: functional assessment staging tool; Depressive symptoms: T-score from the NIH Toolbox PROMIS Emotional Distress – Depression Short Form 4a; QUICKI: quantitative insulin sensitivity check index; Baseline global CBF: baseline global absolute cerebral blood flow in milliliters of blood per 100 grams of tissue per minute (ml/100 g/min).

Clinical characteristics

Clinical blood tests were conducted on all participants at baseline, 3, 6, and 12 months.¹² Additionally, height and weight were collected at baseline and 12 months. The FAST, also known as the Reisberg Functional Assessment Staging Scale, was assessed by a study clinician.²⁴ Insulin resistance was calculated using the Quantitative Insulin Sensitivity Check Index (QUICKI) at baseline, 3, 6, and 12 months.³⁰ Baseline clinical characteristics are presented in Table 1.

Personalized multi-modal medical management and lifestyle intervention

The PREVENTION pilot study is an in-progress, prospective, 12-month, two-arm, RCT (ClinicalTrials.gov NCT04082611). The PREVENTION intervention was previously described in.¹² Briefly, participants in both treatment arms received personalized, evidence-based, multi-modal lifestyle recommendations for improving their brain health from a study physician. These include detailed education and guidelines for exercise, diet, cognitive stimulation, nutritional supplementation, sleep, stress, and medical management. Recommendations, clinical labs, and nutritional supplement intervention were reviewed, and questions were answered during four quarterly physician visits. The intervention period lasted 12 months, with quarterly assessments conducted throughout. The 3- and 6-month timepoints represented mid-intervention assessments, and the 12-month timepoint reflected end-of-intervention outcomes.

The active arm additionally engaged in coaching to help carry out these recommendations. All nutritional supplements and coaching sessions were provided free of charge. They received seven dietary counseling session with registered dietitian nutritionist (RDN), thirty-three live group cognitive-enhanced, multi-component (aerobic, strength, and neuromotor training) exercise classes (^© FitBrain) via video calls led by a certified personal trainer, and thirteen or more overall lifestyle intervention coaching sessions, also with an RDN.

Adherence

Based on interviews and observations during medical appointments, adherence was assessed using a clinician rating scale (CRS)³¹ on an ordinal scale from 1 to 7. A score of 7 indicates active participation, acceptance of the regimen, and some responsibility for managing it; a score of 4 denotes occasional reluctance, such as questioning the need for treatment once a week; and a score of 1 represents complete refusal. Patients with a score ≤ 4 are considered low-adherent.³¹ Although the CRS is ordinal, we also treated it as continuous in some models, consistent with established guidance supporting the use of 5+ point ordinal scales as interval data in parametric analyses.³²

Quantitative magnetic resonance imaging

Brain MRI scans were done on a 3 T General Electric Discovery MR 750 Scanner at baseline and 12 months. Perfusion imaging was performed using a background suppressed 3D pseudo-continuous arterial spin labeling (pCASL) sequence. Scan parameters were as follows: repetition time, 4854 ms; echo time, 10.7 ms; post-labeling delay, 2025 ms; labeling duration, 1450 ms; flip angle, 111°; field of view, 24 cm; image voxel size, 1.875 × 1.875 × 4 mm; acquisition time, 4.30 min. pCASL acquisition included M0 images for reference magnetization in CBF quantification.³³ sMRI was performed using a T1-weighted fast spoiled gradient echo sequence. Scan parameters were as follows: repetition time, 7.91 ms; echo time, 2.96 ms; flip angle, 8°; field of view, 240 × 240; image voxel size, 1 × 1 × 1 mm; acquisition time, 2.53 min.

Arterial spin labeling processing

Absolute CBF was estimated using ASL-MRICloud³⁴ for all pCASL images. Resulting images were converted from radiological to neurological orientation using the FMRIB Software Library (FSL) Linear Image Registration Tool (FLIRT) and registered to the standard Montreal Neurological Institute template.

Statistical analysis

All statistical analyses were conducted using SAS. Prior to analyses, data were inspected for outliers and homogeneity of variance to ensure appropriateness of parametric statistical tests. Treatment arms were compared on demographics and clinical characteristics including CRS and global baseline CBF using a two-tailed two-sample t-test or a χ² test, as appropriate. Completers and drop-outs were also similarly compared. We also examined associations between baseline global CBF and age/sex. The effects of treatment arm and adherence on outcomes (global post-trial CBF and QUICKI) were assessed using a linear mixed effects model, with treatment arm, CRS, time and interaction terms of treatment arm and CRS with time, controlling for age. CRS was considered both as a continuous variable and categorized with a cutoff of ≤ 4 for low adherence. Adherence-based analyses were included to account for individual variability in behaviorally driven treatment exposure. This approach complements traditional group-level comparisons and aligns with theoretical arguments that individualized, non-randomized analyses may yield more informative inferences in small samples.³⁵ We also examined the 3-way interaction of adherence and treatment arm with time. However, this term was not significant for either global CBF (F (2,33) = 0.44, p = 0.6) or QUICKI (F (4,94) = 1.80, p = 0.14), and we therefore report the results without retaining this term in the model. Finally, we considered depressive symptoms and educational level (as proxy for cognitive reserve) as additional covariates; this did not change any of the findings. Significance was set at alpha = 0.05, uncorrected.

We used FSL Randomise to evaluate the effect of treatment arm and adherence on 12-month CBF: a general linear model was estimated, which included treatment arm, CRS, baseline CBF, and age. An alpha level of 0.05 and threshold-free cluster enhancement correction for multiple comparisons was employed.

Results

Participants

Sixty-one patients (mean age 73.4 ± 7.2) were randomized into two groups and followed for one year. Table 1 presents the participant demographics, baseline characteristics, and baseline global CBF. We found no significant or trend differences between treatment arms in demographics or QUICKI measures (all p > 0.1). The control arm had significantly higher baseline global CBF (mean = 27.2 ± 6.7) than the active arm (mean = 22.8 ± 5.4; mean difference = 4.4 ± 6.0; p < 0.01). Baseline global CBF was not significantly associated with age (r = −0.15, p = 0.3) or sex (t (48) = 0.74, p = 0.5).

Of these 61 participants, at the time of data freeze for this study, we had 1-year follow-up data for 21 active and 20 control participants. We had 3-month follow-up data for 21 active and 16 control participants and 6-month follow-up data for 16 active and 15 control participants. Figure 1 presents a visual representation of participant enrollment, randomization, and follow-up. Completers and drop-outs did not differ significantly in any of the baseline characteristics. There was no significant effect of treatment arm on adherence (control: 4.95 ± 1.53; active: 4.55 ± 1.82, t (39) = 0.96, p = 0.3). Adherence was not significantly associated with age (r = 0.05, p = 0.8) or sex (t (39) = 0.80, p = 0.4).

Figure 1.

CONSORT style flow diagram.

Quantitative image analysis

There was a statistically significant effect of treatment arm (see Figure 2, Table 2) and adherence (see Figure 3, Table 3) on post-trial CBF within hypothesized regions, controlling for baseline CBF and age. There was no statistically significant effect of treatment arm or adherence on global CBF.

Figure 2.

Active arm has greater post-trial cerebral blood flow than control arm. Results from two independent sample, two-tailed t-test comparing Active – Control treatment arm post-trial global absolute cerebral blood flow (CBF), accounting for adherence, baseline global CBF, and age. Alpha = 0.05, CWP < 0.05. TOP: Green arrows point to highlighted anatomical regions. Warmer colors represent voxels where active arm had significantly greater post-trial absolute CBF than control arm. BOTTOM: Model residual plots from highlighted significant voxels. Model included baseline global CBF, age, and adherence.

Figure 3.

Adherence to intervention is associated with post-trial cerebral blood flow. Linear relationship between adherence measured using the clinician rating scale and post-trial global absolute cerebral blood flow (CBF) accounting for treatment arm, baseline global CBF, and age. Alpha = 0.05, CWP < 0.05. TOP: Green arrows point to highlighted anatomical regions. Warmer colors represent voxels where more adherent participants had significantly greater post-trial CBF than those with less adherent participants. BOTTOM: Model residual plots from highlighted significant voxels. Model included baseline global CBF, age, and treatment arm.

Table 2.

Treatment arm differences in adjusted post-trial cerebral blood flow.

Region	Right Hemisphere				Left Hemisphere
	x	y	z	p	x	y	z	p
Amygdala	22	−2	−16	0.010*	−80	−6	−14	<0.001*
Angular Gyrus	50	−54	22	0.112	−48	−57	46	0.162
Anterior Cingulate Gyrus	2	24	20	<0.001*	−2	14	30	0.010*
Entorhinal Cortex	14	−3	−23	0.010*	−16	−4	−24	<0.001
Hippocampus	20	−16	−17	<0.001*	−25	−20	−17	0.020*
Inferior Orbitofrontal Gyrus (medial)	15	11	−22	0.182	−20	15	−22	0.010*
Inferior Temporal Gyrus	46	−16	−36	<0.001*	−51	−18	−37	<0.001
Superior Frontal Gyrus (medial)	4	32	53	0.051	−2	28	49	0.051
Middle Frontal Gyrus	52	30	31	0.031*	−30	0	53	0.092
Middle Occipital Gyrus	16	−64	2	0.031*	−10	−71	3	0.182
Middle Temporal Gyrus	57	2	−24	0.132	−62	−23	−25	0.091
Parahippocampal Gyrus	18	−10	−34	<0.001*	−22	−8	−37	0.010*
Posterior Cingulate Gyrus	4	−48	25	0.222	−5	−49	25	0.253
Precentral Gyrus	28	−16	70	0.010*	−28	−22	72	0.020*
Precuneus	20	−57	20	0.010*	−20	−56	14	0.030*
Superior Frontal Gyrus (lateral)	24	12	62	0.152	−24	14	56	0.111
Superior Parietal Gyrus	24	−46	70	0.020*	−23	−46	74	0.023*

Post-trial absolute cerebral blood flow (CBF) accounting for treatment arm, baseline global CBF, and age was analyzed in N = 38 participants. Coordinates are given in Montreal Neurological Institute (MNI) system. *Denotes statistical significance using p < 0.05, corrected for multiple comparisons using threshold free cluster enhancement. Brain regions are included if we expected a difference between treatment arms based on prior literature linking them to: (1) Physical activity – medial temporal lobe (amygdala, hippocampus, and parahippocampal gyrus), posterior cingulate, precuneus, precentral gyrus, and medial superior frontal gyrus. (2) Insulin sensitivity – anterior cingulate, orbitofrontal gyrus, medial and middle frontal gyri, inferior parietal lobule (including angular gyrus), precuneus, posterior cingulate, middle occipital gyrus, and superior frontal and parietal cortices. (3) Cardiovascular risk – anterior cingulate, medial and middle frontal gyri, precentral gyrus, middle temporal gyrus, medial superior frontal gyrus, and lateral temporal cortex

Table 3.

Association between adherence and adjusted post-trial cerebral blood flow.

	Right Hemisphere				Left Hemisphere
Region	x	y	z	p	x	y	z	p
Amygdala	24	−10	−12	0.020*	−20	−6	−20	<0.001*
Angular Gyrus	48	−52	16	0.020*	−42	−52	23	0.010*
Anterior Cingulate Gyrus	5	30	26	0.010*	−2	20	27	<0.001*
Entorhinal Cortex	20	−13	−22	0.031*	−17	−13	−25	0.010*
Hippocampus	24	−12	−20	0.020*	−26	−12	−20	<0.001*
Inferior Orbitofrontal Gyrus (medial)	25	16	−18	0.081	−20	12	−22	<0.001*
Inferior Temporal Gyrus	44	−40	−21	0.010*	−50	−40	−22	0.010*
Superior Frontal Gyrus (medial)	2	16	52	0.020*	−1	15	55	0.010*
Middle Frontal Gyrus	28	72	57	0.020*	−36	18	52	<0.001*
Middle Occipital Gyrus	18	−68	4	0.200	−16	−67	−4	0.011*
Middle Temporal Gyrus	51	−17	−8	0.022*	−52	−29	−9	<0.001*
Parahippocampal Gyrus	32	−31	−18	<0.001*	−16	−10	−24	0.010*
Posterior Cingulate Gyrus	8	−48	23	0.010*	−6	−30	29	<0.001*
Precentral Gyrus	30	−15	49	0.020*	−46	−2	50	0.010*
Precuneus	16	−62	26	<0.001*	−18	−57	11	0.010*
Superior Frontal Gyrus (lateral)	23	1	55	0.081	−14	−2	62	0.010*
Superior Parietal Gyrus	16	−57	71	0.010*	−36	−40	40	<0.001*

Post-trial absolute CBF was analyzed in N = 38 participants and accounted for treatment arm, baseline CBF, and age. Coordinates are given in Montreal Neurological Institute (MNI) system. *Denotes statistical significance using p < 0.05, corrected for multiple comparisons using threshold free cluster enhancement. Brain regions are included if we expected a difference between treatment arms based on prior literature linking them to: (1) Physical activity – medial temporal lobe (amygdala, hippocampus, and parahippocampal gyrus), posterior cingulate, precuneus, precentral gyrus, and medial superior frontal gyrus. (2) Insulin sensitivity – anterior cingulate, orbitofrontal gyrus, medial and middle frontal gyri, inferior parietal lobule (including angular gyrus), precuneus, posterior cingulate, middle occipital gyrus, and superior frontal and parietal cortices. (3) Cardiovascular risk – anterior cingulate, medial and middle frontal gyri, precentral gyrus, middle temporal gyrus, medial superior frontal gyrus, and lateral temporal cortex

Diabetic risk factor

Changes in QUICKI scores were not significant within treatment arms and did not differ between treatment arms (Table 4). However, higher adherence scores were associated with greater improvement in QUICKI, both with adherence treated as a continuous variable (F (3,37) = 3.16, p = 0.04) and using a cutoff score of 4 (F (1,55) = 6.19, p = 0.02) (see Table 4).

Table 4.

Modifiable diabetic risk at baseline and 1-year follow-up.

	ActiveMean (SE)	Within-Group Statistics ^a	Control Mean (SE)	Within-Group Statistics ^a	Between-Group Statistics ^a
Baseline	0.360 (0.007)	t(55) = 0.24, p = 0.8	0.362 (0.007)	t(55) = 0.43, p = 0.7	F(1,55) = 0.02, p = 0.9
1-year	0.362 (0.007)		0.366 (0.007)
	Higher CRS^bMean (SE)	Within-Group Statistics ^a	Lower CRS^bMean (SE)	Within-Group Statistics ^a	Between-Group Statistics ^a
Baseline	0.362 (0.007)	t(55) = 2.13, p = 0.04	0.361 (0.006)	t(55) = −1.41, p = 0.2	F(1,55) = 6.19, p = 0.02
1-year	0.379 (0.006)		0.349 (0.007)

All statistics presented are obtained from mixed effects general linear models, including treatment group, adherence, time, and the interactions between time and treatment group and time and adherence as independent variables, controlling for age. Within-group statistics examine change of the risk measures from baseline to 1-year follow-up and between-group statistics compare the groups on these changes. Diabetic risk was assessed using the quantitative insulin sensitivity check index (QUICKI).

Higher CRS is defined as CRS > 4 and lower CRS is defined as ≤ 4.

Discussion

In this mid-trial analysis, we found evidence that a personalized, multi-modal intervention targeting modifiable cardiovascular and diabetic risk factors through lifestyle modification support and regular medical management increased CBF in patients with early-stage AD symptoms with amyloid-positive pathology.²⁶ We tested whether participants randomized into the PREVENTION Trial active arm or those with greater adherence to the intervention, regardless of treatment arm, exhibited improved CBF, cardiovascular risk, and diabetic risk. In line with our hypotheses, we found that participants randomized into the active arm, when compared with the control arm, had higher post-trial CBF, controlling for baseline global CBF and accounting for adherence and confounders within the inferior orbitofrontal gyrus, MTL, precuneus, the inferior temporal, anterior cingulate, middle frontal, middle occipital, superior parietal, and precentral gyri. The medial superior frontal gyrus was borderline significant. Further, we found that the level of adherence, accounting for treatment arm and confounders, was associated with significantly greater post-trial CBF within the MTL, precuneus, the inferior and middle temporal, anterior and posterior cingulate, precentral gyrus, superior and middle frontal gyrus, superior parietal gyrus, middle occipital, and angular gyrus (part of the inferior parietal lobule). We expected increases in these regions due to their prior links to physical activity, cardiovascular risk, diabetic risk, and relevance to AD.^16–20,36 Taken together, these results suggest that the observed CBF increases may reflect improved cerebrovascular function across multiple, complementary pathways targeted by the intervention. Increased CBF within regions associated with physical activity may also reflect use-dependent plasticity, a process where repeated behaviors increase neuronal activity, which in turn strengthens regional vascular support.³⁷

Many of the regions showing increased CBF in response to the PREVENTION Trial intervention play critical roles in cognitive, sensory, and motor processes affected in AD^38,39 The MTL is critical for memory processing⁴⁰ and is one of the first areas to show atrophy in AD,⁴¹ making it a key target for AD interventions. The anterior cingulate is involved in attention, decision-making, emotional regulation, and memory,⁴² contributing several dementia symptoms including agitation, impaired mood, and hallucinations in patients with AD.⁴³ The precuneus, along with the posterior cingulate are part of the default mode network, which is linked to self-referential thinking and episodic memory, both of which are affected early in AD.⁴⁴ The orbitofrontal gyrus is involved in decision-making, emotion regulation, and social behavior,⁴⁵ domains commonly impacted in early AD. The middle and medial superior frontal gyri are associated with executive function, self-awareness, and goal-directed behavior.⁴⁶ The angular gyrus is implicated in language processing, semantic integration, and episodic memory, and may contribute to both language and memory deficits in AD.⁴⁷

Previous work by our team demonstrated that history of eating a carbohydrate restricted diet was associated with less atrophy within both motor and visual regions in a cross-sectional analysis of patients with early AD.⁴⁸ Motor decline is an early symptom of AD-pathology⁴⁹ and atrophy within the precentral gyrus is associated with poorer motor performance in patients with MCI and AD.⁵⁰ Visual and spatial functions are often subtly impaired in early AD, and several of the regions where we observed increased CBF, including the inferior temporal, middle occipital, and superior parietal cortices, were previously shown to exhibit early hypoperfusion and metabolic decline in neuroimaging studies].³⁸ Taken together with our earlier findings on carbohydrate intake and cortical atrophy, these results may provide complementary evidence across structural and perfusion-based imaging for the impact of diabetic risk management on the amyloid positive brains of patient with early AD symptoms.

Collectively, the locations of our findings suggest that the observed CBF improvements in these regions may hold potential relevance for cognitive and motor function. ASL can detect early alterations in CBF,⁵¹ which often precede structural brain changes, such as atrophy,¹⁵ which in turn lead to cognitive decline.^52,53 ASL is a valuable tool for detecting early vascular changes and assessing the efficacy of interventions for maintaining or improving brain function. However, in AD, CBF and cerebral metabolism may become uncoupled in later stages of the disease due to neurovascular dysfunction, meaning that greater blood flow is needed to deliver the same amount of glucose.⁵⁴ This raises the possibility that increased CBF does not always reflect improved neuronal or metabolic function. In our study, we enrolled individuals with early-stage disease symptoms, where there is likely partial coupling,⁵⁵ and observed regional CBF increases and improved insulin sensitivity (QUICKI) associated with higher adherence, but without direct assessment of cerebral glucose metabolism, we cannot determine with certainty that these vascular and metabolic effects were linked or arose through distinct mechanisms. We emphasize that these findings need to be replicated, ideally with a measure of cerebral glucose metabolism. It will also be essential to examine the correlation of these improvements with the primary outcomes of cognition and hippocampal structure once the study is complete.

Other investigators have also found that lifestyle interventions influence CBF. For example, Guadagni et al. (2020) reported improvements in CBF and cognitive function in older adults following a 6-month aerobic exercise regimen.⁵⁶ Castellano et al. (2017) also utilized ASL to examine the effects of aerobic exercise in patients with early AD, finding increased global glucose metabolism, which correlated with increased CBF.⁵⁷ Notably, Espeland et al. (2018) found a positive impact of a multi-modal lifestyle intervention on CBF in patients with type 2 diabetes.⁵⁸ Our study is the first to find a significant effect of a multi-modal lifestyle and medical management intervention on CBF in patients with early AD clinical symptoms and confirmed amyloidosis.

We also found that greater adherence to the PREVENTION Trial intervention was associated with a significant reduction in diabetic risk over one year. One of the key recommendations of this intervention was that participants were asked to restrict carbohydrates to less than 130 g net carbohydrates per day¹² to increase insulin sensitivity, thereby reducing insulin levels and improving cellular metabolism. Brain cells rely on the insulin signaling through insulin receptors to transport and metabolize glucose.⁵⁹ Cerebral glucose hypometabolism, an early sign of AD-related etiology, is linked to insulin resistance,⁶⁰ which increases with age and is prevalent in AD.⁶¹ AD and type 2 diabetes have shared, underlying molecular mechanisms such as impaired insulin signaling and mitochondrial dysfunction.⁶¹ Both are strongly associated with cognitive decline⁶¹ and lead to amyloid-β (Aβ) and neurofibrillary tangle formation.⁶² Lowering blood glucose and insulin levels may also facilitate the clearance of Aβ peptides,⁶³ which form plaques that disrupt brain cell signaling. Our findings provide preliminary evidence that adherence to the PREVENTION intervention improves insulin sensitivity, a systemic marker of metabolic function. Although we did not evaluate cerebral glucose metabolism, improved systemic insulin sensitivity may have broader implications for cerebrovascular or metabolic health.^64–66 While our randomized group-level analysis did not reveal statistically significant improvements in insulin sensitivity (QUICKI), we did observe significant improvements in participants with higher adherence to the intervention. These findings suggest that improvements in diabetic risk may depend on sufficient engagement with the intervention. Although the active arm received structured coaching and support, adherence scores did not significantly differ between arms, possibly because the active arm was expected to complete a more intensive version of the intervention, while some control participants independently sought external coaching. Though correlation is not causality, this longitudinal finding demonstrates a relationship across multiple time points, where changes in behavior preceded improvements in insulin sensitivity.

Improved systemic glucose metabolism could plausibly be associated with changes in CBF. CBF is generally sensitive to AD-related brain changes^14,15 and often correlates with glucose hypometabolism, as measured by fluorodeoxyglucose-positron emission tomography,⁶⁷ particularly at early stages of the disease.⁵⁵ However, this relationship may weaken as AD progresses due to neurovascular dysfunction.⁵⁴ Because our sample included individuals with early symptoms, it is reasonable to interpret observed CBF increases as potential signs of improved brain function.⁵⁵ Glucose hypometabolism is a widely accepted biomarker of neuronal injury or dysfunction, and is included in the current ATN criteria for diagnosing AD based on biological markers.²⁶ Nonetheless, we interpret CBF changes cautiously, given that glucose metabolism was not assessed and that this is a mid-trial analysis in which the study's primary cognitive outcomes have not yet been analyzed.

The results of the PREVENTION Trial highlight the feasibility of improving cerebrovascular health, as measured with ASL, through lifestyle with standard medical management for AD, cardiovascular, and diabetic risk. Observed ASL results may reflect influences on both vascular disease and AD pathology, particularly in early disease stages where cerebrovascular–metabolic coupling may be at least partially preserved.⁵⁵ Therefore, in real-world contexts, a multi-modal, health-centered approach that supports cerebrovascular function, may confer benefits for brain aging through both vascular and neurodegenerative pathways.^1,4 However, further research is needed to directly link these effects to long-term cognitive outcomes. Interventions targeting modifiable risk factors, such as those in the PREVENTION, FINGER, POINTER, COCOA, and LIEAD trials,^8–12 may offer complementary benefits by addressing both symptom management and underlying cerebrovascular disease, potentially enhancing the effectiveness of newer pharmacological treatments and overcoming barriers to their success.

Limitations of this mid-trial analysis include small sample size and incomplete data collection. This may have impacted our ability to find hypothesized effects of treatment arm on diabetic risk. The small sample size may also limit the effectiveness of randomization. With fewer participants, there is a greater risk that key baseline traits are unevenly distributed across treatment arms despite random assignment, which can confound interpretation of group-level effects.³⁵ We were also unable to detect an interaction between adherence and treatment arm, which again may be due to the small sample size. While we excluded individuals with clinical presentations suggestive of non-AD pathology, we did not collect CSF or tau PET data, limiting our ability to rule out mixed or non-AD neurodegenerative pathologies. Another limitation of this study is the lack of racial, ethnic, and educational diversity. Together with small sample size and the exploratory nature of our outcomes, these limitations reduce the generalizability of the present findings. Finally, although our findings suggest potential physiological benefits of the intervention, cerebral glucose metabolism and cognitive outcomes were not assessed in this analysis. This limits interpretation of the observed CBF changes.

Future directions of this work include analysis of the primary outcomes of the PREVENTION Trial, specifically cognition and hippocampal volume, and examining how these relate to both metabolic and CBF changes reported in this study. Upon trial completion, we will conduct analyses aimed to identify which components of the intervention were most effective. Samples to conduct metabolomics, proteomics, and microbiome analyses and a systems-biology approach were collected as part of this study. These will be examined to gain a deeper understanding of how the intervention impacted additional physiological systems and to inform the development of a future precision-medicine intervention strategy. Replication of findings in larger and more diverse cohorts would also improve generalizability of these exploratory findings. Additional future directions include testing this optimized, PREVENTION intervention in combination with pharmacological treatments to explore potential additive or synergistic benefits for patients at risk for, or diagnosed with, AD.

Footnotes

Acknowledgements

We would like to thank the patients and families that participated in this research, without which this research would not be possible. We thank the Pacific Neuroscience Institute and Foundation staff and leadership, including the CEO and Founder, Dan Kelly, MD, Chief Operating Officer, Chris Cosgrove, Director of Marketing, Zara Jethani, and Executive Assistant, Alyssa Simpers for their support. This study has benefited from the clinical infrastructure of the Pacific Brain Health Center (PBHC), specifically Veronica Bourne, and all the PBHC clinical staff members. We are also grateful to the clinical research infrastructure and leadership provided by the Providence Saint John's, specifically Kim Smith, Jessica Serna, Carrie Mirzna, Lisa van Kreunigen, and Elena Berezhnikh. We are also incredibly thankful to our registered dietitian nutritionists Elizabeth Baron Cole, BS, RDN, CHFS, Evette Richardson, MS, RDN, CPT, and Jordan Stachel, MS, RDN, CPT for dedicating so much time to coordinating care with the PBHC team on behalf of the participants.

ORCID iDs

Jennifer E Bramen

Emily S Popa

Gavin T Kress

Molly K Rapozo

John F Hodes

Ryan M Glatt

Kyron Pierce

David A Merrill

Ethical considerations

This study was approved by the WIRB-Copernicus Group Institutional Review Board (Protocol #20190583) and conducted in accordance with the Declaration of Helsinki (1975).

Consent to participate

All participants provided written informed consent prior to enrollment in the study. For participants at Functional Assessment Staging Tool (FAST) stage 4, a caregiver or legally appointed representative also provided consent and assisted with required study procedures.

Consent for publication

This manuscript does not contain any individual person's data in any form (including individual details, images, or videos). Consent for publication is therefore not applicable.

Author contribution(s)

Jennifer E Bramen: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Resources; Supervision; Visualization; Writing – original draft; Writing – review & editing.

Prabha Siddarth: Formal analysis; Investigation; Methodology; Supervision; Writing – original draft; Writing – review & editing.

Emily S Popa: Data curation; Formal analysis; Writing – review & editing.

Gavin T Kress: Data curation; Formal analysis; Writing – review & editing.

Molly K Rapozo: Project administration; Resources; Writing – review & editing.

John F Hodes: Project administration; Resources; Writing – review & editing.

Aarthi S Ganapathi: Project administration; Writing – review & editing.

William M Sparks: Project administration; Writing – review & editing.

Ynez M Tongson: Project administration; Writing – review & editing.

Andrea M Torres: Project administration; Writing – review & editing.

Somayeh Meysami: Project administration; Writing – review & editing.

Colby B Slyapich: Project administration; Writing – review & editing.

Ryan M Glatt: Conceptualization; Project administration; Resources; Writing – review & editing.

Kyron Pierce: Project administration; Writing – review & editing.

Karen J Miller: Methodology; Supervision; Writing – review & editing.

Shannel H Elhelou: Project administration; Writing – review & editing.

Verna R Porter: Investigation; Writing – review & editing.

Claudia Wong: Project administration; Resources; Writing – review & editing.

Mihae Kim: Project administration; Resources; Writing – review & editing.

Stella Panos: Methodology; Supervision; Writing – review & editing.

Daniel A Hirsch: Resources; Software; Writing – review & editing.

Cyrus A Raji: Investigation; Methodology; Writing – original draft; Writing – review & editing.

Susan Y Bookheimer: Methodology; Writing – review & editing.

Leroy Hood: Conceptualization; Funding acquisition; Methodology; Writing – review & editing.

Jared C Roach: Conceptualization; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Validation; Writing – review & editing.

David A Merrill: Conceptualization; Funding acquisition; Investigation; Methodology; Project administration; Resources; Supervision; Writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work is supported by the Pacific Neuroscience Institute Foundation, including the generous support of the Barbara and John McLoughlin Family as well as the Cary and Will Singleton Family; Providence St Joseph Health, Seattle, WA [Alzheimer's Translational Pillar]; Saint John's Health Center Foundation. MRICloud is supported in part by funding from NIH/NIBIB grant P41EB015909.

Declaration of conflicting interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Cyrus Raji, MD, PhD, is an Editorial Board Member of this journal but was not involved in the peer-review process of this article nor had access to any information regarding its peer-review. The remaining authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data supporting the findings of this study are available on request from the corresponding author.

References

Livingston

Huntley

Liu

, et al. Dementia prevention, intervention, and care: 2024 report of the lancet standing commission. Lancet 2024; 404: 572–628.

van Dyck

Swanson

Aisen

, et al. Lecanemab in early Alzheimer's disease. N Engl J Med 2023; 388: 9–21.

Sims

Zimmer

Evans

, et al. Donanemab in early symptomatic Alzheimer disease: the TRAILBLAZER-ALZ 2 randomized clinical trial. JAMA 2023; 330: 512–527.

Brockmann

Nixon

Love

, et al. Impacts of FDA approval and medicare restriction on antiamyloid therapies for Alzheimer’s disease: patient outcomes, healthcare costs, and drug development. Lancet Reg Health Am 2023; 20: 100467.

NIH National Institute on Aging (NIA). How Is Alzheimer’s Disease Treated?, https://www.nia.nih.gov/health/alzheimers-treatment/how-alzheimers-disease-treated (2023, accessed 18 September 2024).

Doran

Sawyer

. Risk factors in developing amyloid related imaging abnormalities (ARIA) and clinical implications. Front Neurosci 2024; 18: 1326784.

Ngandu

Lehtisalo

Solomon

, et al. A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): a randomised controlled trial. Lancet 2015; 385: 2255–2263.

Roach

Hara

Fridman

, et al. The coaching for cognition in Alzheimer’s (COCOA) trial: study design. Alzheimers Dement (N Y) 2022; 8: e12318.

Ornish

Madison

Kivipelto

, et al. Effects of intensive lifestyle changes on the progression of mild cognitive impairment or early dementia due to Alzheimer’s disease: a randomized, controlled clinical trial. Alzheimers Res Ther 2024; 16: 122.

10.

Kivipelto

Mangialasche

Snyder

, et al. World-Wide FINGERS network: a global approach to risk reduction and prevention of dementia. Alzheimers Dement 2020; 16: 1078–1094.

11.

Baker

Snyder

Espeland

, et al. Study design and methods: U.S. study to protect brain health through lifestyle intervention to reduce risk (U.S. POINTER). Alzheimers Dement 2024; 20: 769–782.

12.

McEwen

Merrill

Bramen

, et al. A systems-biology clinical trial of a personalized multimodal lifestyle intervention for early Alzheimer’s disease. Alzheimers Dement (N Y) 2021; 7: e12191.

13.

Morris

Tangney

Wang

, et al. MIND Diet slows cognitive decline with aging. Alzheimers Dement 2015; 11: 1015–1022.

14.

Iturria-Medina

Sotero

Toussaint

, et al. Early role of vascular dysregulation on late-onset Alzheimer’s disease based on multifactorial data-driven analysis. Nat Commun 2016; 7: 11934.

15.

Zonneveld

Loehrer

Hofman

, et al. The bidirectional association between reduced cerebral blood flow and brain atrophy in the general population. J Cereb Blood Flow Metab 2015; 35: 1882.

16.

Hoscheidt

Kellawan

Berman

, et al. Insulin resistance is associated with lower arterial blood flow and reduced cortical perfusion in cognitively asymptomatic middle-aged adults. J Cereb Blood Flow Metab 2017; 37: 2249–2261.

17.

Hafdi

Mutsaerts

Petr

, et al. Atherosclerotic risk is associated with cerebral perfusion – A cross-sectional study using arterial spin labeling MRI. Neuroimage Clin 2022; 36: 103142.

18.

Moore

Hallsworth

Jakovljevic

, et al. Effects of community exercise therapy on metabolic, brain, physical, and cognitive function following stroke: a randomized controlled pilot trial. Neurorehabil Neural Repair 2015; 29: 623–635.

19.

Yang

Zhuang

, et al. Evaluation of whole-brain oxygen metabolism in Alzheimer’s disease using QSM and quantitative BOLD. Neuroimage 2023; 282: 120381.

20.

Chen

Wolk

Reddin

, et al. Voxel-level comparison of arterial spin-labeled perfusion MRI and FDG-PET in Alzheimer disease. Neurology 2011; 77: 1977–1985.

21.

Kleinloog

JPD

Mensink

Ivanov

, et al. Aerobic exercise training improves cerebral blood flow and executive function: a randomized, controlled cross-over trial in sedentary older men. Front Aging Neurosci 2019; 11: 493764.

22.

Tomoto

Verma

Kostroske

, et al. One-year aerobic exercise increases cerebral blood flow in cognitively normal older adults. J Cereb Blood Flow Metab 2023; 43: 404–418.

23.

Adhikari

Hong

Zhao

, et al. Cerebral blood flow and cardiovascular risk effects on resting brain regional homogeneity. Neuroimage 2022; 262: 119555.

24.

Reisberg

. Functional assessment staging (FAST). Psychopharmacol Bull 1988; 24: 653–659.

25.

Sjostrom

Ainsworth

Bauman

, et al. Guidelines for data processing analysis of the International Physical Activity Questionnaire (IPAQ) – Short and long forms, https://www.researchgate.net/file.PostFileLoader.html?id=5641f4c36143250eac8b45b7&assetKey=AS%3A294237418606593%401447163075131 (2005, accessed 26 April 2023).

26.

Jack

Andrews

Beach

, et al. Revised criteria for diagnosis and staging of Alzheimer’s disease: Alzheimer’s association workgroup. Alzheimers Dement 2024; 20: 5143–5169.

27.

Johnson

Minoshima

Bohnen

, et al. Update on appropriate use criteria for amyloid PET imaging: dementia experts, mild cognitive impairment, and education. J Nucl Med 2013; 54: 1011–1013.

28.

Schindler

Gray

Gordon

, et al. Cerebrospinal fluid biomarkers measured by Elecsys assays compared to amyloid imaging. Alzheimers Dement 2018; 14: 1460–1469.

29.

Meyer

Kirmess

Eastwood

, et al. Clinical validation of the PrecivityAD2 blood test: a mass spectrometry-based test with algorithm combining %p-tau217 and Aβ42/40 ratio to identify presence of brain amyloid. Alzheimers Dement 2024; 20: 3179–3192.

30.

Vanhala

Kumpusalo

, et al. The quantitative insulin sensitivity check index QUICKI predicts the onset of type 2 diabetes better than fasting plasma insulin in obese subjects: a 5-year follow-up study. J Clin Endocrinol Metab 2002; 87: 5834–5837.

31.

Kemp

David

. Psychological predictors of insight and compliance in psychotic patients. Br J Psychiatry 1996; 169: 444–450.

32.

Sullivan

Artino

. Analyzing and interpreting data from Likert-type scales. J Grad Med Educ 2013; 5: 541–542.

33.

Clement

Petr

Dijsselhof

MBJ

, et al. A beginner’s guide to arterial spin labeling (ASL) image processing. Frontiers in radiology. Front Radiol 2022; 2: 929533.

34.

Liu

, et al. ASL-MRICloud: an online tool for the processing of ASL MRI data. NMR Biomed 2019; 32: e4051.

35.

Jaynes

. Probability theory: the logic of science. Cambridge: Cambridge University Press, 2003.

36.

Kapasouri

Ioannidis

Cameron

, et al. The utility of arterial spin labeling MRI in medial temporal lobe as a vascular biomarker in Alzheimer’s disease spectrum: a systematic review and meta-analysis. Diagnostics (Basel) 2022; 12: 2967.

37.

Bogorad

DeStefano

Linville

, et al. Cerebrovascular plasticity: processes that lead to changes in the architecture of brain microvessels. J Cereb Blood Flow Metab 2019; 39: 1413.

38.

Talwar

Kushwaha

Chaturvedi

, et al. Systematic review of different neuroimaging correlates in mild cognitive impairment and Alzheimer’s disease. Clin Neuroradiol 2021; 31: 953–967.

39.

Scheltens

Blennow

Breteler

MMB

, et al. Alzheimer’s disease. Lancet 2016; 388: 505–517.

40.

Cutsuridis

Yoshida

. Editorial: memory processes in medial temporal lobe: experimental, theoretical and computational approaches. Front Syst Neurosci 2017; 11: 259179.

41.

Thompson

Hayashi

De Zubicaray

, et al. Dynamics of gray matter loss in Alzheimer’s disease. J Neurosci 2003; 23: 994–1005.

42.

Rolls

. The cingulate cortex and limbic systems for emotion, action, and memory. Brain Struct Funct 2019; 224: 3001.

43.

Chen

Dang

Zhang

. Brain mechanisms underlying neuropsychiatric symptoms in Alzheimer’s disease: a systematic review of symptom-general and –specific lesion patterns. Mol Neurodegener 2021; 16: 38.

44.

Buckner

Andrews-Hanna

Schacter

. The brain’s default network: anatomy, function, and relevance to disease. Ann N Y Acad Sci 2008; 1124: 1–38.

45.

Rolls

. The orbitofrontal cortex and emotion in health and disease, including depression. Neuropsychologia 2019; 128: 14–43.

46.

Friedman

Robbins

TW.

The role of prefrontal cortex in cognitive control and executive function. Neuropsychopharmacology 2022; 47: 72–89.

47.

Seghier

. The angular gyrus: multiple functions and multiple subdivisions. Neuroscientist 2013; 19: 43.

48.

Bramen

Siddarth

Popa

, et al. Impact of eating a carbohydrate-restricted diet on cortical atrophy in a cross-section of amyloid positive patients with Alzheimer’s disease: a small sample study. J Alzheimers Dis 2023; 96: 329–342.

49.

Buchman

Bennett

. Loss of motor function in preclinical Alzheimer’s disease. Expert Rev Neurother 2011; 11: 665.

50.

Koppelmans

Silvester

Duff

. Neural mechanisms of motor dysfunction in mild cognitive impairment and Alzheimer’s disease: a systematic review. J Alzheimers Dis Rep 2022; 6: 307.

51.

Wierenga

Hays

Zlatar

ZZ.

Cerebral blood flow measured by arterial spin labeling MRI as a preclinical marker of Alzheimer’s disease. J Alzheimers Dis 2014; 42(Suppl 4): S411–S419.

52.

Coupé

Manjón

Lanuza

, et al. Lifespan changes of the human brain in Alzheimer’s disease. Sci Rep 2019; 9: 3998.

53.

Bakkour

Morris

Dickerson

. The cortical signature of prodromal AD: regional thinning predicts mild AD dementia. Neurology 2009; 72: 1048–1055.

54.

Kisler

Nelson

Montagne

, et al. Cerebral blood flow regulation and neurovascular dysfunction in Alzheimer disease. Nat Rev Neurosci 2017; 18: 419–434.

55.

Alsop

Dai

Grossman

, et al. Arterial spin labeling blood flow MRI: its role in the early characterization of Alzheimer’s disease. J Alzheimers Dis 2010; 20: 871–880.

56.

Guadagni

Drogos

Tyndall

, et al. Aerobic exercise improves cognition and cerebrovascular regulation in older adults. Neurology 2020; 94: E2245–E2257.

57.

Castellano

Paquet

Dionne

, et al. A 3-month aerobic training program improves brain energy metabolism in mild Alzheimer’s disease: preliminary results from a neuroimaging study. J Alzheimers Dis 2017; 56: 1459–1468.

58.

Espeland

Luchsinger

Neiberg

, et al. Long term effect of intensive lifestyle intervention on cerebral blood flow. J Am Geriatr Soc 2018; 66: 120–126.

59.

Kahn

. Type 2 diabetes: when insulin secretion fails to compensate for insulin resistance. Cell 1998; 92: 593–596.

60.

Willette

Bendlin

Starks

, et al. Association of insulin resistance with cerebral glucose uptake in late middle-aged adults at risk for Alzheimer’s disease. JAMA Neurol 2015; 72: 1013.

61.

Michailidis

Moraitou

Tata

, et al. Alzheimer’s disease as type 3 diabetes: common pathophysiological mechanisms between Alzheimer’s disease and type 2 diabetes. Int J Mol Sci 2022; 23: 2687.

62.

Rad

Arya

Karimian

, et al. Mechanism involved in insulin resistance via accumulation of β-amyloid and neurofibrillary tangles: link between type 2 diabetes and Alzheimer’s disease. Drug Des Devel Ther 2018; 12: 3999.

63.

Mullins

Diehl

Chia

, et al. Insulin resistance as a link between amyloid-beta and tau pathologies in Alzheimer’s disease. Front Aging Neurosci 2017; 9: 118.

64.

Deery

Liang

Di Paolo

, et al. Peripheral insulin resistance attenuates cerebral glucose metabolism and impairs working memory in healthy adults. NPJ Metab Health Dis 2024; 2: 17.

65.

Kodl

Seaquist

. Cognitive dysfunction and diabetes mellitus. Endocr Rev 2008; 29: 494–511.

66.

Kullmann

Heni

Hallschmid

, et al. Brain insulin resistance at the crossroads of metabolic and cognitive disorders in humans. Physiol Rev 2016; 96: 1169–1209.

67.

Dolui

Nasrallah

, et al. Arterial spin labeling versus 18F-FDG-PET to identify mild cognitive impairment. Neuroimage Clin 2020; 25: 102146.