APOE ε4 and Dietary Patterns in Relation to Cognitive Function: An Umbrella Review of Systematic Reviews

Abstract

Context

Carrying the apolipoprotein ε4 allele (APOE ε4) is the strongest genetic risk factor for late-onset Alzheimer’s disease. There is some evidence suggesting that APOE ε4 may modulate the influence of diet on cognitive function.

Objective

This umbrella review of systematic reviews evaluates the existing literature on the effect of dietary interventions on cognitive and brain-imaging outcomes by APOE status.

Data Sources

PubMed, EMBASE, Web of Science, and Scopus were searched using terms appropriate to each area of research, from their respective starting dates of coverage until March 2023.

Data Extraction

Two independent reviewers conducted data extraction and performed a quality appraisal using the Measurement Tool to Assess Systematic Reviews (AMSTAR) 2

Data Analysis

Six total reviews were included in the final analysis. Four reviews evaluated randomized controlled trials on individuals aged 50–93 years ranging the entire cognitive continuum. One review combined observational studies and clinical trials conducted on both cognitively healthy and cognitively impaired individuals (age range: 50–90), and 1 review included observational studies of both cognitively healthy and cognitively impaired adults (age range: 50–75).

Results

Both observational studies and clinical trials yielded inconclusive results attributed to both practical limitations associated with longitudinal follow-up and issues of methodological quality. Except for the Mediterranean diet, dietary interventions, such as the ketogenic diet, nutraceuticals, and supplements, were generally not effective in older APOE ε4 carriers. This review considers plausible biological mechanisms that might explain why older and cognitively impaired APOE ε4 carriers were less likely to benefit.

Conclusion

This review identifies notable gaps in the literature, such as a shortage of studies conducted in middle-aged and cognitively healthy APOE ε4 carriers assessing the impact of dietary interventions and provides suggestions for novel trial designs.

INTRODUCTION

The role of diet in the treatment or prevention of Alzheimer’s disease (AD) remains unclear as evidence from observational studies and clinical trials has been inconclusive and sometimes conflicting.¹ Observational studies suggest that diets rich in saturated fats are associated with a higher risk of cognitive decline, while greater consumption of both monounsaturated and polyunsaturated fatty acids (PUFAs) is associated with a lower risk of age-related cognitive decline.² Diets that are commonly consumed by Western populations increase the risk of age-related cognitive impairment due to their high saturated fat content and are associated with neuroinflammation later in life.³ However, several randomized clinical trials (RCTs) testing the effect of nutritional interventions on cognitive outcomes largely report null effects, with the active intervention arm performing similarly to the control arm.¹ Nevertheless, the use of dietary supplements and nutraceuticals to promote cognitive health among older adults is becoming increasingly more prevalent, despite a lack of compelling evidence and targeted recommendations guiding their use. Therefore, careful evaluation of the existing literature on nutritional interventions and their potentially protective effects against brain aging and cognitive decline is essential.

The brain’s response to dietary interventions appears to be modulated by genetic risk. For example, a genetic variant can modify the effect of diet on the course of AD. Conversely, diet and the epigenetic changes it induces can modify the expression of genes. The apolipoprotein ε4 allele (APOE ε4), which is the strongest genetic risk factor for developing late-onset AD,⁴ has been shown to modify the effects of diet on the brain. However, previous studies have reported conflicting results.⁵ Some studies found that APOE ε4 carriers adhering to healthy diets and lifestyles are generally protected from cognitive decline, while others show no effect of dietary and lifestyle patterns on cognitive outcomes.^4,6  APOE is heavily involved in lipid transport and influences the metabolism of nutrients in the brain across the lifespan.⁷ Of relevance to the diet, the brains of APOE ε4 carriers exhibit a decline in blood–brain barrier (BBB) functions and changes to mitochondrial functions that impact both brain-nutrient transport and nutrient metabolism.⁸ Since BBB function and mitochondrial functions decline with age and disease,⁷ older APOE ε4 carriers are less likely to benefit from dietary interventions.

Several systematic reviews examined the quality of diet studies and cognition but have not addressed the effects of APOE ε4. Umbrella reviews offer a systematic and efficient approach to synthesize evidence from multiple systematic reviews, facilitating evidence-based decision-making, and guiding future research directions. Therefore, this umbrella review aims to present and evaluate published systematic reviews related to diet, APOE ε4, and cognition and to identify knowledge gaps, discuss potential mechanisms for the observed findings, and suggest appropriate future research directions.

METHODS

Protocol and Registration

Before any analyses were conducted, this review’s protocol was registered in Open Science Framework (OSF; https://doi.org/10.17605/OSF.IO/WHF87). Ethics approval was not required as the review did not involve original experimentation.

Literature Search

The search strategy was developed by the librarian (J.E.D.) and included the following databases on March 28, 2023: PubMed/MEDLINE (National Center for Biotechnology Information; coverage: 1946–present), Embase (Embase and Embase Classic, Elsevier; coverage: 1947–present), Web of Science (Web of Science Core Collection, Clarivate Analytics; coverage: 1900–present), and Scopus (Elsevier; coverage: 1970–present). The full search strategy for each database can be found in the Supplementary Appendix. All citations were imported into Covidence (Covidence.org) for deduplication and subsequent screening.

Search Strategies for Inclusion

The search strategy was designed to identify systematic reviews and meta-analyses on the effects of diet and nutritional supplementation in cognitively normal, older adults or individuals with AD dementia or cognitive impairment. Specifically, this search sought out systematic reviews that investigated the effects of the Mediterranean diet, ketogenic diet, fatty acid supplementation, and other nutritional interventions and cognition. The systematic reviews that this search yielded were considered for further evaluation regardless of whether or not they performed a meta-analysis. A list of key search terms can be found in the Supplementary Appendix. APOE ε4 was not included in the search criteria as its inclusion drastically reduced the number of reviews.

Data Synthesis

No data synthesis was performed.

Assessment of Overlap

Overlapping studies were identified and duplicate searches were removed.

Methodological Quality

The methodological quality of these studies was evaluated using the Measurement Tool to Assess Systematic Reviews (AMSTAR) 2.

Risk of Bias

Two independent reviewers conducted the AMSTAR 2 review to reduce bias.

RESULTS

This review was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for study selection (Figure 1). A preliminary search identified a total of 98 publications. Following deduplication, the number of publications to be evaluated for eligibility was reduced to 62 reviews that underwent title and abstract screening. A total of 32 reviews failed to meet eligibility criteria (see Table S1 for full list of excluded publications and justification for exclusion) and of the remaining 30 reviews that underwent full-text review, 22 studies were excluded (see Table S1), leaving a total of 8 systematic reviews that met the inclusion criteria. Papers that were (1) not systematic reviews, (2) not relevant to the research topic, and (3) not addressing and analyzing APOE ε4 and dietary factors to a detailed extent were excluded.

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Flow Chart of the Study Selection and Inclusion Process. A sum of 98 records were initially identified, 6 of which are analyzed in the following review

Two independent reviewers (T.J.U., J.P.) assessed the 8 selected articles with AMSTAR 2, which is an online tool used to conduct a critical appraisal of systematic reviews and meta-analyses.⁹ The 2 reviewers assessed the articles independently and a third reviewer (A.A.T.) was consulted to reach consensus on any disagreements.

Two reviews that met fewer than 2 AMSTAR 2 criteria for methodological quality were excluded from further analysis and discussion.

Characteristics of Included Systematic Reviews

The types of dietary patterns and interventions examined ranged considerably across the included systematic reviews. Two publications focused on the effect of ketogenic interventions,^10,12 1 study focused on n-3 fatty acids and the Mediterranean diet,¹³ 1 study focused on saturated and trans-fat intake,¹⁴ and 2 studies examined multiple dietary patterns and nutraceuticals.^15,16 Outcomes of these reviews generally focused on the intervention effects on cognition based on cognitive testing, blood biomarkers, and brain structure based on neuroimaging. The Population, Intervention, Comparison, Outcome, Study design (PICOS) criteria used to define the research question are presented in Table 1. Details about each systematic review are included in Table 2.^10–16 Primary studies of ketogenic interventions identified in the systemic reviews when the APOE ε4 effect was described are also included.

Table 1.

PICOS Criteria for Inclusion of Studies

Parameter	Criterion
Population	Older adults ≥50 years
Interventions	Nutritional interventions (eg, Mediterranean diet, ketogenic diet, nutraceuticals)
Comparisons	Placebo, no intervention, or other dietary pattern
Outcomes	Cognition, brain function
Study design	Systematic reviews and meta-analyses of randomized controlled trials (RCTs) and observational studies

Table 2.

Key Summary of the Included Reviews and Their Pertinent Outcomes and Findings

Study, year	Number of databases searched	Population studied; age range of participants	Study types included	Number of studies included; combined sample size	Intervention/exposure variables	Outcomes; effect	APOE ε4 effect (carriers and noncarriers)
Grammatikopoulou et al, 202010	3 (+ gray literature)	Patients with MCI or AD diagnosis; 50–93 y	Randomized controlled trials	10; 456	Ketogenic therapy (diet/supplementation)	Global cognition testing; positive effect	Delayed responsiveness to ketogenic benefits in APOE ε4
Masana et al, 201713	3	Cognitively healthy; >65 y	Observational studies and randomized controlled trials	24; 47 963	n–3 Fatty acids, adherence to the Mediterranean diet	Global cognitive testing, MRI assessments; positive effects on cognition, mixed effects on brain volumes	No central conclusion due to limited APOE ε4 investigation
Castro et al, 202312	5	Individuals with subjective cognitive decline, MCI, or AD; 65+ y	Randomized controlled trials	17; 1067	Supplementation with medium-chain fatty acids	Serum analysis and global cognitive testing; positive effect on blood ketones, mixed effect on cognition	No conclusive data suggesting benefits of supplementation; benefits of supplementation found
D'Cunha et al, 201815	4	Subjects aged over 65 y; 65+ y	Randomized controlled trials	25; 22 744	Nutraceuticals and dietary supplements	Global cognitive assessments, activities of daily living testing; inconclusive effect	APOE ε4 carriers less likely to benefit from nutritional interventions
Barnard et al, 201414	3	Older adults; 50–75 y	Observational studies	12; 22 226	Saturated and trans-fat intake	Incident dementia, AD, or MCI or cognitive decline; greater risk of cognitive decline	Higher risk of AD for APOE ε4 carriers with high saturated fat intake; no measurable risk for non APOE ε4 carriers
Hersi et al, 201716	14	Older adults; 50–90 y	Clinical trials and observational studies	9; Not reported	Polyunsaturated fatty acids; vitamins B12, B6, C, E, and B3/niacin; folic acid/folate; multivitamins; flavonoids; fruit and vegetable intake; beta-carotene; caffeine consumption; Mediterranean diet; fat intake; and saturated fatty acids	Incident AD dementia; inconclusive	Insufficient data on APOE ε4 subgroups

Abbreviations: AD, Alzheimer’s disease; APOE ε4, apolipoprotein ε4 allele; fMRI, functional magnetic resonance imaging; MCI, mild cognitive impairment; MRI, magnetic resonance imaging; PET, positron emission tomography.

Quality of Included Systematic Reviews

A critical appraisal of the quality of each systematic review was conducted using the AMSTAR 2 checklist.⁹ All included studies had at least 1 AMSTAR 2 item that was inadequately addressed, indicating that none of the included systematic reviews are void of limitations (Figure 2). Two reviews failed to address either all or all but 1 item on the AMSTAR 2 checklist^11,17 and were therefore excluded from further discussion of findings due to their particularly poor methodological quality, yielding a total of 6 systematic reviews to be further discussed. A schematic representation of the methodological quality of the included studies can be seen in Figure 2.

Figure 2.

Detailed Overview of AMSTAR 2 Scoring for Each of the 8 Extracted Articles. Abbreviation: PICO, Population, Intervention, Comparison, Outcome

Ketogenic Interventions

Two of the included publications focused on the effects of ketogenic interventions, including supplementation with medium-chain fatty acids. Grammatikopoulou et al¹⁰ reviewed 10 RCTs with samples ranging from 6 to 152 subjects diagnosed with either mild cognitive impairment (MCI) or AD, totaling 456 pooled subjects. Participants included in this review were between 50 and 93 years of age. Most of the trials reviewed reported a dampened effect of ketogenic interventions on cognitive outcomes among APOE ε4 carriers, compared with noncarriers with MCI or AD. However, several of those studies were likely not sufficiently powered to detect differences between APOE ε4 subgroups. Furthermore, 1 study demonstrated that APOE ε4 carriers exhibited a delayed response to the ketogenic intervention, suggesting that the absence of therapeutic benefits in this group might be due to the relatively short duration of these studies (45–180 days). This review adequately addressed most AMSTAR 2 items but did not justify all exclusion criteria for their search (eg, language), did not search the gray literature or the reference lists of included studies, and did not provide a list of excluded publications and reasons for exclusion (Figure 2).

The second systematic review¹² on the effect of ketogenic interventions on cognition and circulating ketones reviewed 17 clinical trials conducted on patients diagnosed with MCI or AD, with a combined sample of 1067 subjects aged 65 and older. Of these 17 clinical trials, a total of 13 studies measured cognitive outcomes in response to a ketogenic intervention. Eight of the clinical trials that assessed cognition outcomes accounted for participant APOE genotype. Three trials reported that the intervention significantly improved cognition only for APOE ε4 noncarriers. One RCT demonstrated that a ketone intervention increased regional cerebral blood flow only among APOE ε4 noncarriers. Interestingly, these studies demonstrated similar and sometimes even exaggerated increases in blood ketones among APOE ε4 carriers compared with noncarriers, suggesting that the lack of therapeutic effects on cognition is not due to a lack of absorption or systematic metabolism.

To provide deeper insights on how APOE ε4 effect may influence the response of a ketogenic diet on the brain, primary RCTs from the systematic reviews that included APOE ε4 were examined. The RCTs reviewed by Grammatikopoulou et al¹⁰ and Castro et al¹² examining the effect of ketogenic supplementation on cognition by APOE status are summarized in Table 3,^18–25 and illustrate the differences in cognitive responses to ketogenic interventions, stratified by APOE ε4 status. Two major RCTs did not find an effect of ketogenic supplementation on cognition, independent of APOE status.^18,21 Henderson et al²¹ reported no significant APOE ε4 effect on cognitive functioning, which the authors attributed to a lack of ketogenic formulation bioavailability coupled with a high rate of participant withdrawal. Fortier et al¹⁸ was inadequately powered to detect an APOE ε4 effect on cognitive outcomes or ketosis, due to an insufficient sample size.

Table 3.

Summary of APOE ε4 Effect on Ketogenic Supplementation in Randomized Controlled Trials

Study, year	Population	Sample size, n	APOE ε4+	Duration	Intervention	Primary outcome(s)	Secondary outcome(s)	APOE ε4 effect
Fortier et al, 201918	MCI	52	14	6 mo	MCT 30 g/d vs placebo	Δ in CMR	Cognitive tests	Ketone metabolism enhancement in the treatment group. No effect of APOE ε4 on cognitive outcomes or ketosis.  Note: Insufficient sample size to determine APOE ε4 effect.
Fortier et al, 202119	MCI	82	19	6 mo	MCT 30 g/d vs placebo	Cognitive tests	Δ in CMR-A; Δ in plasma ketone	In a subanalysis of APOE ε4+ placebo vs kMCT scores on several cognitive tests (RL/RL16 total recall, Stroop, Boston Naming Testa) did not change. In an APOE ε4 subanalysis, there was a significant difference of cognitive scores between placebo vs kMCT groups for these outcomes.
Henderson et al, 200920	MTM AD	152	69	104 d	MCT AC-1202 10 g/d = 7 d; then, 20 g/d for remainder vs placebo	Cognitive tests (Δ in ADAS-Cog and ADCS-CGIC)	MMSE, ß-OHB	Treatment enhanced cognitive benefit in APOE ε4–negative patients and correlated with serum ß-OHB levels.
Henderson et al, 202021	MTM AD	413	128	6 mo	MCT AC-1204 increase 10 g/every fourth day for 2 wk, then 20 g/d for remainder vs placebo	Cognitive tests (Δ in ADAS-Cog11)	Cognitive tests, ß-OHB	Treatment did not enhance cognitive benefit across all patients. Note: Null results attributed to insufficient formulation absorption.
Rebello et al, 201522	MCI	4	1	6 mo	MCT 56 g/d vs placebo	Cognitive tests	ß-OHB	More enhanced memory and ketone uptake in APOE ε4 noncarriers vs APOE ε4 carriers. ß-OHB enhancement in APOE ε4 carriers.
Reger et al, 200423	AD, MCI	20	10	2 h	Emulsified MCT 40 mL vs placebo	Cognitive tests	ß-OHB	Treatment enhanced cognitive performance in APOE ε4–negative patients.
Torosyan et al, 201824	MTM AD	16	8	45 d	MCT 40 g/d vs placebo	Cerebral blood flow imaging analysis	Volumetric change	Enhanced rCBF observed in APOE ε4–negative patients.
Xu et al, 202025	MTM AD	53	3	30 d	MCT 17.3 g/d vs placebo	Cognitive tests	ADL score	Cognitive benefit observed among APOE ε4–negative patients.

^aBoston Naming Test = language measure.

Abbreviations: AD, Alzheimer’s disease; ADAS-Cog, Alzheimer’s Disease Assessment Scale-Cognitive Subscale; ADCS-CGIC, Alzheimer’s Disease Cooperative Study–Clinical Global Impression of Change; ADL, activities of daily living scale; APOE ε4, apolipoprotein ε4 allele; ß-OHB, Beta-hydroxybutyrate; CMR, cerebral metabolic rate; kMCT, ketogenic medium chain triglyceride; MCI, mild cognitive impairment; MCT, medium-chain triglycerides; MMSE, Mini-Mental State Examination; MTM, mild-to-moderate; rCBF, regional cerebral blood flow; RCT, randomized controlled trial; RL/RL16, 16-item free/cued word learning recall test, episodic memory measure; Stroop, Stroop color-word interference test, executive function measure.

Among trials that reported an APOE effect on cognition in response to ketogenic supplementation, the largest trial contained 152 participants with mild-to-moderate AD, 69 of whom were ε4 positive.²⁰ The effects of ketogenic agents on cognition differed by APOE ε4 status. Unlike noncarriers, APOE ε4 carriers did not show any response on Alzheimer’s Disease Assessment Scale-Cognitive Subscale (ADAS-Cog) after 45 and 90 days of medium-chain triglyceride (MCT) supplementation, despite increased ketone β-OHB (beta-hydroxybutyrate) concentrations in plasma.²⁰ Similarly, Reger et al²³ reported significant β-OHB increases 90 minutes following MCT ingestion in the treatment condition among APOE ε4 carriers. However, APOE ε4 noncarriers had significant improvements in ADAS-Cog scores, whereas APOE ε4 carriers displayed a slight decrease in cognitive performance. In a post hoc analysis, another trial reported improvements in executive function among APOE ε4 noncarriers compared with placebo, while no effect was observed among APOE ε4 carriers.¹⁹ The study was not adequately powered to detect an APOE ε4 by treatment interaction effect.

Polyunsaturated Fatty Acids

Three of the included reviews examined the relationship between PUFA intake and cognition in older adults.^13,15,16 The systematic review by Masana et al¹³ reviewed a total of 24 observational studies and clinical trials that included participants aged 65 years and older, but concluded that there was insufficient evidence to support a moderating effect of APOE ε4. The second review by Hersi et al¹⁶ examined the effect of PUFAs on adults between the ages of 50 and 90 years and included primary studies that did not conduct any APOE ε4 subgroup analyses. D’Cunha et al¹⁵ reviewed 3 placebo-controlled RCTs with a combined sample of 4180 participants aged 65 and older investigating the effect of PUFA supplementation on cognitive performance in healthy older adults, patients with a history of myocardial infarction, and patients with AD. Of the 3 trials, only the AD trial stratified analyses by APOE ε4 status and reported that omega-3 fatty acid supplementation resulted in slower rates of cognitive decline only among APOE ε4–negative patients with AD.

Mediterranean Diet

The review by Masana et al¹³ examined the relationship between the Mediterranean diet and cognitive decline among cognitively intact older adults aged 65 years and older. The review included 1 clinical trial and 7 observational studies. The influence of APOE ε4 was examined only in the clinical trial and demonstrated no differences in the effect of extra-virgin olive oil supplementation on cognition in APOE ε4 carriers and noncarriers. However, it should be noted that the authors only tested for an interaction effect and did not perform stratified analyses. Given the small proportion of APOE ε4 carriers in each experimental group, their sample may have lacked the statistical power to detect an interaction effect.

Hersi et al¹⁶ reported 4 included systematic reviews and 1 primary study that examined the association between adherence to the Mediterranean diet and AD risk. Of the 4 included systematic reviews, 1 study in an included systematic review found evidence of a protective effect. The study included APOE ε4 status as a covariate but did not stratify the sample and perform the analyses in APOE ε4 carriers and noncarriers separately. Another included systematic review in Hersi et al identified 1 study that indicated an association between the Mediterranean diet and reduced risk of AD, and a third included systematic review identified 3 studies as such. The included primary study reported a reduction in AD risk associated with increased Mediterranean diet compliance.

Saturated and Trans-Fat Intake

Barnard et al¹⁴ reviewed a total of 12 observational studies with adults aged 50–75 years examining the association of consuming saturated and trans-fat on incident AD or other dementia, MCI, or cognitive decline among cognitively healthy older adults. Among the 4 prospective studies that focused on incident AD or other forms of dementia as their outcome, 3 performed APOE ε4 subgroup analyses; 2 of these 3 studies found that associations between total energy intake, total fat intake, or saturated fat intake and incident AD were driven primarily by APOE ε4 carriers. Of the 4 prospective studies with incident MCI as an outcome, only 1 performed APOE ε4 subgroup analyses; this study found that saturated fat intake was associated with a greater risk of MCI only among APOE ε4 carriers. The authors concluded that APOE ε4 carriers may have a higher risk of AD dementia and MCI associated with higher intake of saturated and trans-fat. The review did not account for any potential risk of bias in the included studies (Figure 2).

Nutraceuticals and Dietary Supplements

Hersi et al¹⁶ reviewed the effect of several other dietary factors and supplements (including vitamin B₁₂, vitamin B₆, vitamin B₃/niacin, folic acid/folate, vitamins C and E, multivitamins, flavonoids, fruit and vegetable intake, beta-carotene, and caffeine consumption) on incident AD and related dementias, but the review did not report on any APOE ε4 subgroup analyses associated with nutraceuticals and dietary supplements. Similarly, D’Cunha et al¹⁵ investigated the impact of other dietary patterns and supplements, including herbal supplements and vitamin B supplements, and combinations of nutraceuticals on cognitive outcomes and AD risk; the review did not report APOE ε4 subgroup analyses.

DISCUSSION

The systematic reviews included in this umbrella review provide moderate evidence for certain dietary interventions having a protective effect on cognitive performance in older adults. The majority of the systematic reviews identified did not demonstrate similar protective effects of dietary interventions in APOE ε4 carriers, with the exception of Hersi et al,¹⁶ who reported that the relationship between Mediterranean diet and cognitive outcomes was not affected by APOE ε4 status. Nevertheless, the study they reviewed did not formally test for an interaction between APOE ε4 and the Mediterranean diet on the outcomes of interest. D’Cunha et al¹⁵ revealed that APOE ε4 carriers were generally less likely to benefit from ketogenic interventions. Finally, Barnard et al¹⁴ demonstrated that APOE ε4 carriers are more likely to be at risk of developing AD if their diet is high in saturated fats.

How Does Aging and APOE ε4 Affect the Response to Dietary Intervention?

Some observational studies provide evidence that a healthy dietary pattern, such as that of the Mediterranean diet in older APOE ε4 carriers, is less effective.⁵ However, these same dietary patterns have been associated with less cognitive disease or lower AD risk in younger APOE ε4 carriers.²⁶ As noted in most of the included reviews here, many participants were of a quite advanced age and some had MCI or AD, conditions that can lower the ability of nutrients to enter the brain or alter their metabolic fates.⁷ As AD pathology advances, compromised BBB integrity or mitochondrial dysfunction directly impact nutrient brain uptake and metabolism.^27,28 The endothelial cells of the BBB require higher energy demands to fuel nutrient transport systems and maintain BBB integrity.²⁹ Several nutrients, including vitamins B₇, B₉, and B₁₂, the omega-3 fatty acid docosahexaenoic acid (DHA), and vitamin D all have a lower cerebral spinal fluid (CSF) to blood ratio in patients with AD, which can be possibly explained by compromised nutrient-transport function of the BBB or their metabolism in the AD brain.⁷

The APOE ε4 brain cannot metabolize glucose efficiently as neurodegeneration advances, and ketones or fat have been proposed as an alternate source of energy. However, some studies suggest that ketone brain metabolism is also impaired in patients with AD. For example, decreased CSF ketone levels in patients with AD and MCI compared with controls³⁰ indicate either an increase in ketone brain loss (via catabolic/oxidative pathways) or lower transport of ketones across the BBB. Both human and rodent studies have reported BBB dysfunction in association with APOE ɛ4.^31,32  APOE ɛ4–associated BBB damage may impede the transport of important nutrients into the brain, such as ketone bodies. In support of this hypothesis, 1 study found that monocarboxylate transporters responsible for ketone body transport across the BBB are less functional in APOE ε4 carriers.³³ Mitochondrial functions also appear to be dysregulated in the brains of APOE ɛ4 carriers with dementia,³⁴ such as enzyme functions involved in ketone oxidation.³⁵ These findings are also consistent with prior evidence of more reactive oxygen species and lower NAD(+)/NADH ratio associated with disrupted mitochondrial respiration.³⁶ Similarly, omega-3 PUFAs have accelerated brain catabolism in older APOE ε4 carriers with mild dementia,³⁷ with upregulation of phospholipases such as calcium-dependent phospholipase A2 in APOE ε4.³⁸

Maladaptive lipid metabolism may also, in part, account for the dampened response to dietary interventions among APOE ɛ4 carriers. Carrying the APOE ε4 allele is associated with higher concentrations of low-density-lipoprotein (LDL) cholesterol^39,40 and postprandial triglycerides⁴¹ (a surrogate of insulin resistance [IR]), which are exacerbated by saturated fat intake or high-carbohydrate diets enriched with simple sugars, respectively.⁴² Lipid metabolism and IR are worsened with aging, and these metabolic pathways are accentuated in older APOE ɛ4 carriers.

Can Dietary Interventions Influence the Onset and Course of AD in People Who Carry the APOE ε4 Allele?

APOE ε4 carriers exhibit cerebral amyloid-β (Aβ) accumulation at an early age (30s–40s), with a gene–dose effect, such that homozygotes accumulate Aβ earlier than heterozygotes.⁴³ It may be hypothesized that cerebral Aβ accumulation (fibrillar plaques) before the onset of advanced tauopathy (such as with neuritic plaques) and neurodegeneration represents reversible “brain stress” and a window for interventions, as illustrated in Figure 3. Among the different dietary patterns, diets enriched in fibers, polyphenols, and mono- or polyunsaturated fats are associated with improved metabolic profiles (eg, greater insulin sensitivity, improved lipid levels), making such dietary patterns a logical choice for improving cognitive functions in those carrying the APOE ε4 allele and with increased risk of dementia.⁴ Thus, targeting younger and cognitively normal APOE ε4 carriers with a Mediterranean or Mediterranean-DASH (-Dietary Approaches to Stop Hypertension) intervention for neurodegenerative delay (MIND)–like dietary pattern may yield more favorable outcomes, potentially delaying the onset of clinical AD than if started later in life, particularly when clinical disease is present. Newer trial designs are needed to address this hypothesis.

Figure 3.

Mechanisms That Influence the Response to Nutritional Intervention in Both an APOE ε4 Preclinical Brain and a Dementia Brain. Mechanisms that largely impact an APOE ε4 carrier in the preclinical stage include (1) subtle cognitive impairment; (2) reversible biochemical, cellular, and network changes; (3) amyloid accumulation; (4) vulnerability to vascular risk factors. In an APOE ε4 dementia brain, the brain exhibits (1) irreversible neurodegeneration, (2) mitochondrial dysfunction, coupled with (3) disrupted blood–brain barrier function. The implications of these mechanisms on the APOE ε4 brain’s response to nutritional intervention are largely unknown. Nutritional interventions administered during the preclinical stage may therefore be more effective among APOE ε4 carriers compared with interventions administered at a later stage. This figure was created by BioRender. Abbreviations: APOE ε4, apolipoprotein ε4 allele; DASH, Dietary Approaches to Stop Hypertension; MIND, Mediterranean-DASH intervention for neurodegenerative delay

Two Contrasting Trial Designs Can Address the Role of the Diet on the Brain in APOE ε4 During Midlife

The Nutrition for Dementia Prevention Working Group was tasked to develop a future roadmap to improve nutritional trial design for brain outcomes.¹ Two interventional designs were proposed to better inform the effects of diet on the brain and are of relevance to those who carry the APOE ε4 allele. A personalized approach with small sample sizes and surrogate outcomes (eg, blood pressure, gut microbiome, plasma AD biomarkers) could target APOE ε4 carriers with hypertension and metabolic risk factors during middle age, or a population-based approach with practical multimodal interventions and scalable outcomes that leverages advances in digital health (eg, phone-based apps, remote technologies, electronic medical records). Studying subgroups based on increased risk, such as APOE ε4, early in life with either personalized or population interventions holds substantial promise for targeted dementia prevention in high-risk groups.

Biomarker Outcomes as Surrogates of Brain Function for Assessing the Role of the Diet in the Younger APOE ε4 Brain

Cognitive decline in APOE ε4 carriers is often evident at or after 55 years of age,⁴⁴ although more subtle changes may be observed even earlier. Notably, most of the systematic reviews discussed here included studies of adults aged 50 years and older. Given the need for earlier preventive approaches, biomarkers of brain health are important. Fluid biomarkers, neuroimaging techniques, the gut microbiome, along with vascular and metabolic markers can provide useful surrogate measures of brain health earlier in life before the onset of neurodegeneration or cognitive impairment to assess the efficacy of dietary interventions during the preclinical phase.⁴⁵

Neuroimaging techniques, such as magnetic resonance imaging (MRI), have been demonstrated to be effective in the evaluation of subtle structural changes and the prediction of future cognitive changes,⁴⁶ which could prove particularly useful in future studies that investigate nutritional intervention effects in healthy older adults. Advances in MRI technologies, such as magnetic resonance spectroscopy (MRS), phase contrast MR (PC MR), and T2 relaxation under tagging (TRUST MR), enable studying the uptake and metabolism of several nutrients, and here ketones are provided as an example. The quantification of ketogenic supplementation on cerebral metabolism is feasible through MRS. It has been established that induced ketosis elevates brain ketone levels in low millimolar quantities.⁴⁷ Phase contrast MR enables rapid and accurate assessment of cerebral flow to the brain through the 4 great arteries of the neck.^48,49 TRUST MR yields a rapid and noninvasive assessment of venous oxygenation (SVO₂) in the superior sagittal sinus that can capture oxygen extraction as an index for the oxidation of ketones in the brain.^12,50 Additional imaging modalities include ketone positron emission tomography (PET), which may be used to monitor regional ketone uptake with specificity but poor spatial resolution and directly relate such measures to improvements in cognitive outcomes.¹⁹ To this end, studies testing the effect of APOE ε4 on brain ketone uptake at different ages and stages of the cognitive continuum are warranted.

The gut microbiome is another promising biomarker to guide the efficacy of dietary interventions. Altered gut microbiota have also been implicated in AD pathogenesis. Distinct microbiome signatures have been identified among patients with MCI in comparison to their cognitively normal counterparts. Microbial distinctions among patients with MCI correlated positively to CSF markers of AD pathology, such that a positive association between Proteobacteria and the ratio of Aβ was reported. Through a randomized, cross-group dietary intervention in 17 MCI and cognitively normal participants, a modified Mediterranean-ketogenic diet resulted in changes in microbiome populations, such as increased Enterobacteriaceae and decreased Bifidobacterium, which are associated with decreased AD biomarker concentrations among both groups.⁵¹ In another randomized crossover study of 20 participants with MCI or who were cognitively normal who primarily consumed a low-fat American Heart Association diet or a high-fat modified Mediterranean ketogenic diet, decreased γ-aminobutyric acid (GABA)–producing microbiota and elevated amounts of GABA-regulating microbiota were observed among participants who adhered to the modified ketogenic diet. This regulatory effect is repressed by APOE ε4 with elevated concentrations of GABA-producing microbes in the CSF and gut of APOE ε4 carriers.⁴⁹

Dietary modifications have also been associated with changes in cerebral Aβ biomarkers. In 2 cohort longitudinal studies with a follow-up duration of 3 years, increased adherence to a Mediterranean diet was associated with decreased accumulation of cerebral Aβ.^52,53 Clinical trials of a dietary intervention composed of low concentrations of saturated fatty acids along with a low glycemic index has been reported to improve Aβ profile, with observed increases in CSF Aβ₄₂.^54,55 Excessive consumption of saturated fatty acids and dietary components with a high glycemic index has been associated with decreased CSF Aβ₄₂⁵⁵ and elevated amyloid concentrations in the brain.⁵⁶ Plasma AD biomarkers are rapidly emerging as noninvasive alternatives to CSF and PET imaging assessments of Aβ₄₂ status that can help select participants or assess the efficacy of dietary interventions on these biomarkers. McKhann et al⁵⁷ highlighted the utility of biomarker usage in the diagnosis of AD and dementia and noted the specificity of this tool. Further, Atri⁵⁸ emphasized the ability of biomarker-based tests to detect physiological changes in the body well before symptom onset for AD and dementia, which further highlights the need for biomarker diagnostic testing to best treat AD well before dementia symptoms begin to appear.

Vascular and metabolic markers, such as hypertension, IR or dyslipidemia, also provide meaningful surrogate outcomes of relevance to brain functions. Hypertension is an established modifiable dementia risk factor⁵⁹ that has been shown to interact with APOE ε4 to increase AD risk during midlife.⁶⁰ The DASH study investigated the effects of dietary patterns on blood pressure in individuals with diastolic blood pressure between 80–95 mmHg in 459 participants over an 11-week period. Those randomized to the combination diet had a significant decrease in systolic and diastolic blood pressure compared with those on the control diet.⁶¹ Schelke et al⁶² referenced the development of IR, oxidative stress, and dyslipidemia, which are closely associated with AD pathogenesis and how adherence to the MIND dietary pattern, as well as the practice of regular exercise (three 30-min sessions/wk) can be effective. It was also noted that supplementation of omega-3 fatty acids and B vitamins were associated with reduced dyslipidemia and oxidative stress, respectively. Therefore, new trials could test the effects of a similar diet on APOE ε4 carriers with hypertension or metabolic risk factors in midlife, with blood pressure reduction, IR in relation to plasma, or imaging AD biomarkers as the main outcomes.

Strengths and Limitations

One notable strength of this umbrella review is its systematic and comprehensive approach in identifying relevant reviews. However, given that the goal of this umbrella review is to present and evaluate other systematic reviews, the current work is impacted by the limitations of the reviewed studies. Several of the included reviews failed to address multiple aspects of methodological quality on the AMSTAR 2 checklist. One particularly prevalent limitation across studies was the absence of appropriate assessment of bias. Many of the primary studies evaluated by the included reviews were not sufficiently powered to detect APOE differences in the effects of dietary interventions. Furthermore, several studies reported high variability from differences in cognitive outcomes, making it difficult to synthesize the data and perform meta-analysis. It is also important to consider differences in dietary interventions across cultural contexts. Several variations of different diets exist, such as the Mediterranean diet, which incorporate different nutritional components to varying degrees and might complicate comparisons across studies.¹³ The design of clinical trials was variable, with notable protocol and outcome measure differences. This further precludes the ability to conduct an in-depth analysis of the effect of diet on brain health. Both RCTs and observational studies were limited by unaccounted participant adherence rate, high rate of dropout, lack of APOE stratification, and insufficient sample size. Therefore, large RCTs with well-defined populations are needed to further investigate the relationship between APOE status, diet, and brain health.

CONCLUSION

This umbrella review has provided a comprehensive overview of the existing literature regarding the protective effects of dietary interventions against cognitive impairment in older adults. Additionally, important gaps in the literature are identified, such as a shortage of studies investigating the impact of dietary interventions on cognitive and clinical outcomes in APOE ε4 carriers earlier in life. Assessing the efficacy of interventions in APOE ε4 carriers earlier in life might be necessary to circumvent some of the metabolic limitations and pathological processes that would dampen the APOE ε4 brain’s response if administered later in life. The use of neuroimaging techniques and noninvasive biomarkers of AD risk will be vital in the assessment of the efficacy of these interventions in high-risk individuals earlier in the aging process. Carefully designed clinical trials that examine how interventions interact with genetic risk will critically inform precision-medicine recommendations to reduce the disease burden in those who are at highest risk.

Notes

The protocol is registered with Open Science Framework (OSF) registries. Further registration information can be found at https://doi.org/10.17605/OSF.IO/WHF87

Author Contributions

T.J.U., A.A.T., and N.A. contributed equally to this work. H.N.Y. conceptualized and designed the review and extensively edited the manuscript. T.J.U., A.A.T., J.P., W.J.M., V.K.C., and J.E.D. wrote the initial version of the manuscript. T.J.U., A.A.T., and J.P. did the AMSTAR reviews. N.A. significantly contributed to doing the primary literature analysis, created the figure, and contributed to the results and discussion. W.J.M. and V.K.C. reviewed and designed the research methodology. J.E.D. was the librarian designing the search methodology.

Supplementary Material

Supplementary Material is available at Nutrition Reviews online.

Funding

H.N.Y. holds the Kenneth and Bette Volk Endowed Chair of Neurology. H.N.Y. is supported by RF1AG076124, RF1AG078362, R01AG067063, R01AG054434, R01AG055770, R21AG056518, and P30AG066530 from the National Institute on Aging; GC-201711–2014197 from the Alzheimer’s Drug Discovery Foundation (ADDF); and generous donations from the Vranos and Tiny Foundations and from Ms. Lynne Nauss.

Conflicts of Interest

None declared.