Differential Associations of APOEɛ 2 and APOEɛ 4 Genotypes with Cerebrospinal Fluid Biomarkers of Alzheimer’s Disease in Individuals Without Dementia

Abstract

Background:

The APOE genotype has emerged as the major genetic factor for AD but differs among different alleles.

Objective:

To investigate the discrepant effects of APOE genotype on AD cerebrospinal fluid (CSF) biomarkers.

Methods:

A total of 989 non-demented ADNI participants were included. The associations of APOE ɛ2 and APOE ɛ4 with CSF biomarkers were investigated using linear regression models. Interaction and subgroup analyses were used to investigate the effects of sex and age on these associations. Furthermore, we used mediation analyses to assess whether Aβ mediated the associations between APOE genotypes and tau.

Results:

APOE ɛ2 carriers only showed higher Aβ levels (β [95% CI] = 0.07 [0.01, 0.13], p = 0.026). Conversely, APOE ɛ4 carriers exhibited lower Aβ concentration (β [95% CI] = –0.27 [–0.31, –0.24], p < 0.001), higher t-Tau (β [95% CI] = 0.25 [0.08, 0.18], p < 0.001) and higher p-Tau (β [95% CI] = 0.31 [0.25, 0.37], p < 0.001). Subgroup analysis showed that APOE ɛ2 was significantly positively associated with Aβ only in females (β [95% CI] = 0.12 [0.04, 0.21], p = 0.005) and older people (β [95% CI] = 0.06 [0.001, 0.12], p = 0.048). But the effects of APOE ɛ4 were independent of gender and age. Besides, the associations of APOE ɛ4 with t-Tau and p-Tau were both mediated by baseline Aβ.

Conclusions:

Our data suggested that APOE ɛ2 could promote Aβ clearance, while the process could be modified by sex and age. However, APOE ɛ4 might cause the accumulation of Aβ and tau pathology independent of sex and age.

Keywords

Age interaction Alzheimer’s disease amyloid-β sex interaction tau

INTRODUCTION

Alzheimer’s disease (AD) is the leading cause of dementia in the elderly. AD is characterized by the presence of amyloid-β (Aβ) plaques and neurofibrillary tangles which are thought to be responsible for neurodegeneration and subsequent cognitive dysfunction [1]. Due to a lack of effective treatments for AD, a better understanding of AD-related genetic and demographic factors is of great value in clinical practice.

The apolipoprotein E (APOE) allele is strongly correlated with AD. The three most common genotypes of the APOE gene are ɛ2, ɛ3, and ɛ4, among which the ɛ3 allele is the most common and is often used as the reference to evaluate the risk of AD. Current evidence suggests that APOE ɛ4 is associated with the accumulation of Aβ [2 –5] and tau neurofibrillary tangles [6 –10]. Contrary to the detrimental effects of APOE ɛ4 on AD, the APOE ɛ2 is protective against AD [3 , 12]. However, few studies have been conducted on APOE ɛ2 because of its low prevalence in populations (∼8% in general populations and ∼5% in AD populations) [13, 14]. Although previous cerebrospinal fluid (CSF) analyses and PET studies have reported the relationship between APOE ɛ2 and AD pathology, their results were inconsistent [15, 16], likely due to the small sample sizes [15], different effects of population characteristics (e.g., sex and age) and methodological differences [16]. Besides, the question whether APOE genotypes impacts tau pathology in an Aβ dependent pathway remains controversial[17 –21].

In addition to genetic susceptibilities, sex and age are both important demographic factors that influence AD risk. When it comes to sex, the females have been reported to have higher incidence of AD [22], increased AD pathology and worse cognition, compared to males [23, 24]. Moreover, a critical but often overlooked feature of the APOE genotype-AD link is that it may be more pronounced in women [25]. A previous multisite study based on the results in CSF biomarkers and neuropathology showed that APOE ɛ4 conferred a greater AD risk in women [26]. However, whether sex modifies the influence of APOE ɛ2 on AD remains unknown. In terms of age, a recent large multicenter study showed that age modified the association between APOE ɛ4 and AD. The impact of APOE ɛ4 is most pronounced between 65 and 70 years. However, the protective association between APOE ɛ2 genotype and the risk of AD might not be influenced by age [27]. Although sex and age differences have been well established in a comprehensive meta-analysis [25], very little is known about the underlying mechanisms. Taken together, the interaction between demographic factors and APOE genotype remains controversial and is an area of intense investigation inthe future.

Our study aimed to investigate the different influence of APOE ɛ2 and ɛ4 genotypes on AD biomarkers in the subgroups stratified by sex and age in individuals without dementia, which will provide implications for clinical AD prevention and diagnosis.

MATERIALS AND METHODS

ADNI study design

ADNI is a multi-site open access dataset designed to accelerate the discovery of biomarkers to identify and track AD pathology (https://adni.loni.usc.edu/). Data collection and sharing in ADNI were approved by the Institutional Review Board of each participating institution, and written informed consent was obtained from all participants.

Participants

From the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database, we selected all participants without dementia (cognitively normal [CN] and mild cognitive impairment [MCI]). The MCI participants had objective memory loss tested by the delayed recalled of the Weschersler Memory Scale logical memory II, scored 24 or higher on baseline Mini-Mental State Examination (MMSE) [28], scored 0.5 on the Clinical Dementia Rating (CDR) scale [29], preserved activities of daily living, and absence of dementia. The CN participants had MMSE scores of 24 or higher and a CDR global score of 0. Individuals with subjective memory complaints at baseline were included in the CN group [30].

All the included participants should have the following baseline data: diagnoses in the ADNI database, APOE genotypes, as well as CSF levels for Aβ₄₂, t-Tau, and p-Tau (n = 989). Participants were grouped as APOE ɛ2 carriers (ɛ2/ɛ2 or ɛ2/ɛ3, n = 89), APOE ɛ3 homozygotes (n = 498), and APOE ɛ4 carriers (ɛ3/ɛ4 or ɛ4/ɛ4, n = 402). In addition, we excluded APOE ɛ2/ɛ4 participants from the analysis due to the small sample size (n = 13).

APOE genotypes

APOE genotypes were determined by genotyping the two single-nucleotide polymorphisms that define the APOE ɛ2, ɛ3, and ɛ4 genotypes (rs429358, rs7412) with DNA extracted by Cogenics from a 3-mL aliquot of EDTA blood (https://adni.loni.usc.edu/data-samples/genetic-data/). Polymerase chain reaction (PCR) amplification was followed by HhaI restriction enzyme digestion, resolution on 4% Metaphor Gel, and visualization by ethidium bromide staining [31].

CSF measurements

CSF collection and processing were described in detail at https://adni.loni.usc.edu/methods/. Briefly, CSF biomarkers were measured at the ADNI Biomarker Core Laboratory (University of Pennsylvania) using the multiplex xMAP Luminex platform (Luminex Corp, Austin, TX) with Innogenetics (INNO-BIA AlzBio3; Ghent, Belgium; for research use only reagents) immuno-assay kit-based reagents containing 4D7A3 monoclonal antibody for Aβ₄₂, AT120 monoclonal antibody for t-Tau, and AT270 monoclonal antibody for p-Tau. All CSF biomarker assays were performed in duplicate and averaged values were used. Besides, the concentrations were log transformed for analysis.

Neuroimaging measurements

Magnetic resonance imaging (MRI) was performed using a Siemens Trio 3.0 T scanner or Vision 1.5 T scanner (GE, Siemens, and Philips). Regional volume estimates were processed using Free-surfer software package version 4.3 and 5.1 image processing framework for the MRI images, respectively. ROIs included the hippocampus [32 –34]. Estimated intracranial volume (ICV) was used to adjust for head size variation. FDG-PET data were acquired and reconstructed according to a standardized protocol (https://adni.loni.ucla.edu/). For FDG-PET, we averaged the counts of angular, temporal, and posterior cingulate regions.

Neuropsychological assessments

We downloaded the widely used composite scores of episodic memory (ADNI-MEM) and executive function (ADNI-EF) [35] from the ADNI-LONI website, as well as the recently validated composites of language (ADNI-LAN) and visuospatial functioning (ADNI-VS) [36]. Compared with the separate scales, these composite scales can reflect the cognition more comprehensively.

Statistical analysis

We classified participants into three groups according to their APOE genotypes: APOE ɛ2 carriers, APOE ɛ4 carriers, and APOE ɛ3/ɛ3 homozygotes. We compared all available data (demographics, CSF biomarkers, hippocampal volumes, FDG-PET, and cognitive scores) in different APOE groups using Kruskal-Wallis tests for continuous variables and Chi-squared tests for categorical variables, respectively. Holm-Bonferroni correction method was used for multiple intergroups comparisons.

The primary cross-sectional analysis assessed the associations between APOE genotypes and AD phenotypes. We performed this analysis using linear regression with APOE genotypes (APOE ɛ2 carriers, APOE ɛ4 carriers, APOE ɛ3/ɛ3 homozygotes) as an independent variable, and all available characteristics as dependent variables after adjusting for age, sex, education, and diagnosis in independent models (hippocampal volume should be corrected by ICV). For all analyses, APOE ɛ3/ɛ3 participants were selected as the reference group. In addition, all the dependent variables were log transformed for analysis. In the secondary analysis, we examined APOE by sex and age interactions on all the phenotypes. To improve the results, subgroup analyses stratified by age and sex were conducted.

We performed post hoc mediation analyses to test whether the associations between APOE and CSF biomarkers were mediated via Aβ₄₂ pathway. Age, sex, years of education, and baseline diagnosis were included as covariates in these models and this analysis was only performed for the associations that were significant in the primary analysis described above. Besides, we also repeated this mediation analysis in different subgroups stratified by sex and age.

The indirect effect was estimated with the significance determined using 10,000 bootstrapped iterations. The “lme4,” “lmerTest,” “mediation,” and “ggplot2” packages in R 4.1.1 software were used to perform the above analyses. A two-tailed p < 0.05 was considered statistically significant.

RESULTS

Demographic characteristics

The demographic and clinical information of the participants were presented in Table 1. Overall, participants were mainly MCI (62.9% MCI) and were evenly distributed by sex (45.1% women), with a mean age of 72.94 years and average years of education of 16.18. In terms of the demographic differences across APOE groups, APOE ɛ4 carriers were slightly younger than the other two groups (p < 0.001). APOE ɛ4 carriers had a lower proportion of CN compared to APOE ɛ3/ɛ3 homozygotes and APOE ɛ2 carriers (32.7% versus 43.8% versus 56.2%) (p < 0.001).

Table 1

Demographic and clinical information for all participants and subdivided by APOE genotypes

Characteristics	Total	APOEɛ2 carriers	APOEɛ3/ɛ3	APOEɛ4 carriers	p
	(N = 989)	(N = 89)	(N = 498)	(N = 402)
Demographics
Age, y, mean (SD)	72.94 (6.99)	73.36 (6.48)	73.81 (7.04)	71.76 (6.89)	<0.001
Gender, female, N (%)	446 (45.1%)	43 (48.3%)	225 (45.2%)	178 (44.3%)	0.786
Education, y, mean (SD)	16.18 (2.71)	15.94 (2.94)	16.31 (2.63)	16.07 (2.76)	0.3844
Diagnosis, CN, N (%)	367 (37.1%)	50 (56.2%)	218 (43.8%)	99 (32.7%)	<0.001
CSF biomarker^†
Aβ₄₂ (ng/mL)	182.44 (53.48)	216.63 (48.49)	200.15 (49.46)	152.93 (44.45)	<0.001
t-Tau (ng/mL)	267.88 (117.83)	228.22 (98.02)	241.29 (101.47)	310.01 (127.96)	<0.001
p-Tau₁₈₁ (ng/mL)	25.53 (13.15)	20.75 (1.10)	22.33 (10.83)	30.60 (14.63)	<0.001
Structure imaging(volume)^‡
Hippocampus	7086.41 (1084.86)	7263.710526 (1017.41)	7184.59 (1061.13)	6920.64 (1110.61)	0.001 ^¶
PET Imaging^§
FDG PET	1.27 (0.13)	1.30 (0.12)	1.29 (0.13)	1.25 (0.14)	<0.001
Cognition scores
Episodic memory	0.51 (0.76)	0.79 (0.76)	0.64 (0.71)	0.28 (0.76)	<0.001
Executive function	0.44 (0.90)	0.69 (0.93)	0.50 (0.90)	0.31 (0.89)	<0.001
Language	0.43 (0.80)	0.54 (0.74)	0.51 (0.80)	0.32 (0.80)	0.001
Visuospatial functioning	0.03 (0.71)	0.01 (0.72)	0.10 (0.69)	–0.06 (0.72)	0.003

^†CSF biomarkers data were missing for 22 t-Tau, 23 p-Tau. ^‡Hippocampal volume data were missing for 120 hippocampus. ^§PET imaging data were missing for 161 FDG-PET. ^¶Hippocampal volumes were corrected by ICV. SD, standard deviation; CN, cognitively normal; MCI, Mild cognitive impairment; APOE, apolipoprotein E; CSF, cerebrospinal fluid; Aβ, amyloid-β; p-Tau 181, tau phosphorylated at threonine 181; t-Tau, total tau; PET, positron emission tomography-computed tomography; FDG, Fluorodeoxyglucose.

Associations between APOE genotypes and AD phenotypes

We used APOE ɛ3/ɛ3 homozygotes as the reference group, APOE ɛ2 carriers showed higher Aβ₄₂ levels (β [95% CI] = 0.07 [0.01, 0.13], p = 0.026) but no differences in p-Tau and t-Tau (Table 2 and Fig. 1). Conversely, APOE ɛ4 carriers had lower Aβ₄₂ concentration (β [95% CI] = –0.27 [–0.31, –0.24], p < 0.001), higher t-Tau (β [95% CI] = 0.25 [0.08, 0.18], p < 0.001), and higher p-Tau (β [95% CI] = 0.31 [0.25, 0.37], p < 0.001) (Table 2 and Fig. 1). As for the imaging biomarkers, APOE ɛ4 carriers exhibited lower hippocampal volumes (adjusted by ICV) (β [95% CI] = –0.25 [–0.37, –0.13], p < 0.001) (Table 2 and Supplementary Figure 1) and lower FDG PET (β [95% CI] = –0.33 [0.47, –0.19], p < 0.001) (Table 2 and Supplementary Figure 2). But APOE ɛ2 carriers showed no significant differences in both imaging biomarkers. Broadly, regarding cognitive scores, APOE ɛ2 genotype was not significantly associated with any cognitive composite scores, though this genotype was marginally positively associated with MEM (β [95% CI] = 0.06 [–0.11, 0.23], p = 0.082). On the contrary, APOE ɛ4 genotype were significantly negatively correlated with MEM (β [95% CI] = –0.47 [–0.60, –0.35], p < 0.001), EF (β [95% CI] = –0.18 [–0.30, –0.07], p = 0.002), LAN (β [95% CI] = –0.15 [–0.27, –0.04], p = 0.01) and VS (β [95% CI] = –0.18 [–0.31, –0.05], p = 0.01) (Table 2 and Supplementary Figure 3).

Table 2

Linear regression parameters (standardized β) for the association of APOE ɛ2 and APOE ɛ4 genotypes with AD phenotypes

	APOEɛ2 carriers		APOEɛ4 carriers
	β [95% CI]	p	β [95% CI]	p
CSF biomarkers
Aβ₄₂	0.07 [0.01, 0.13]	0.026	–0.27 [–0.31, –0.24]	<0.001
t-Tau	–0.03 [–0.01, 0.01]	0.489	0.25 [0.08, 0.18]	<0.001
p-Tau₁₈₁	–0.04 [–0.14, 0.06]	0.408	0.31 [0.25, 0.37]	<0.001
Structure imaging (corrected ICV)
Hippocampus	–0.05 [–0.25, 0.16]	0.643	–0.25 [–0.37, –0.13]	<0.001
PET Imaging
FDG PET	–0.03 [–0.26, 0.20]	0.800	–0.33 [–0.47, –0.19]	<0.001
Cognition scores
Episodic memory	0.06 [–0.11, 0.23]	0.082	–0.47 [–0.60, –0.35]	<0.001
Executive function	0.13 [–0.07, 0.32]	0.207	–0.18 [–0.30, –0.07]	0.002
Language	–0.04 [–0.24, 0.15]	0.670	–0.15 [–0.27, –0.04]	0.010
Visuospatial functioning	–0.17 [–0.39, 0.05]	0.134	–0.18 [–0.31, –0.05]	0.010

AD phenotypes included CSF biomarkers, hippocampal volume, FDG-PET, and cognition scores. APOE ɛ3/3 participants were selected as the reference group for all comparisons. Models included age, sex, education, and baseline diagnosis as covariates (hippocampus volume was corrected by ICV). Bold indicated that the results were statistically significant (p < 0.05). APOE, apolipoprotein E; CSF, cerebrospinal fluid; Aβ, amyloid-β; p-Tau 181, tau phosphorylated at threonine 181; t-Tau, total tau; PET, positron emission tomography-computed tomography; FDG, Fluorodeoxyglucose; CI, confidence interval.

Fig. 1

Associations of APOE genotypes with CSF biomarkers (A-C). Differences in CSF biomarker levels were examined by Games-Howell test. Each path of the model was corrected for age, sex, years of education and baseline diagnosis. CSF, cerebrospinal fluid; Aβ, amyloid-β; p-Tau, phosphorylated tau protein; t-Tau, total tau protein.

Associations between APOE genotypes and AD phenotypes in subgroup analyses

Interaction of APOE genotypes with sex or age was found in CSF biomarkers (Supplementary Tables 1 and 2). The subgroup analysis stratified by sex indicated that APOE ɛ2 had significantly positive association with Aβ₄₂ in females (β [95% CI] = 0.12 [0.04, 0.21], p = 0.005), but no significant association in males (β [95% CI] = 0.02 [–0.06, 0.10], p = 0.661) (Fig. 2 and Supplementary Table 3). As for APOE ɛ4 carriers, the effects of APOE ɛ4 on CSF biomarkers, imaging results and the cognitive scores were almost independent of gender (Fig. 2 and Supplementary Table 3), except that EF scores (β [95% CI] = –0.25 [–0.43, –0.07], p = 0.007) and VS scores (β [95% CI] = –0.32 [–0.52, –0.12], p = 0.002) were significantly negatively associated with APOE ɛ4 only in women. When it comes to age, the subgroup analysis stratified by age showed that APOE ɛ2 genotype was significantly positively associated with Aβ₄₂ in older people (β [95% CI] = 0.06 [0.001, 0.12], p = 0.048), but not significantly associated with Aβ₄₂ in younger participants. Additionally, APOE ɛ4 genotype was significantly negatively associated with FDG-PET scores (β [95% CI] = –0.03 [–0.05, –0.02], p < 0.001), MEM scores (β [95% CI] = –0.30 [–0.41, –0.19], p < 0.001), EF scores (β [95% CI] = –0.16 [–0.30, –0.03], p = 0.015) and VS scores (β [95% CI] = –0.15 [–0.29, –0.01], p = 0.033) only in older people, while CSF biomarkers and hippocampal volumes were almost independent of age (Fig. 3 and Supplementary Table 4).

Fig. 2

Linear regression parameters for the association between APOE and AD phenotypes in the subgroups stratified by sex. APOE ɛ3/ɛ3 carriers were selected as the reference group for all APOE comparisons. The model included age, education, and diagnosis as covariates (hippocampal volume was also corrected by ICV). 0.01≤*≤0.05, 0.001≤** < 0.01, 0.0001≤*** < 0.001. APOE, apolipoprotein E; CSF, cerebrospinal fluid; Aβ, amyloid-β; p-Tau 181, tau phosphorylated at threonine 181; t-Tau, total tau; PET, positron emission tomography-computed tomography; FDG, Fluorodeoxyglucose.

Fig. 3

Linear regression parameters for the association between APOE and AD phenotypes in the subgroups stratified by age. APOE ɛ3/ɛ3 carriers were selected as the reference group for all APOE comparisons. The model included sex, years of education, and basic diagnosis as covariates (hippocampal volume was also corrected by ICV). 0.01≤*≤0.05, 0.001≤** <0.01, 0.0001≤*** <0.001. APOE, apolipoprotein E; CSF, cerebrospinal fluid; Aβ, amyloid-β; p-Tau 181, tau phosphorylated at threonine 181; t-Tau, total tau; PET, Positron emission tomography-computed tomography; FDG, Fluorodeoxyglucose.

Mediation analysis results

In the mediation analyses, we found that the association between APOE ɛ4 and tau was mediated by Aβ₄₂ (Fig. 4A). Specifically, Aβ₄₂ partially mediated the association between APOE ɛ4 and t-Tau, leaving a significant Aβ₄₂ independent effect of APOE ɛ4 on t-Tau that corresponded to 44.31% of the total effect (Fig. 4A). Besides, Aβ₄₂ partially mediated the association between APOE ɛ4 and p-Tau, leaving a significant Aβ₄₂ independent effect of APOE ɛ4 on p-Tau burden that corresponded to 54.98% of the total effect (Fig. 4A).

Fig. 4

Mediation effects of Aβ pathology on the association of APOE ɛ4 genotypes and tau pathologies. The bold p values indicate the mediation pathways are meaningful. The proportions shown in the figure indicate the proportion of mediating factors in the total effect of amyloid pathology on t-Tau and p-Tau. The model included age, sex, years of education and basic diagnosis as covariates. Aβ, amyloid-β; p-Tau, phosphorylated tau protein; t-Tau, total tau protein.

Sex-stratified mediation analyses revealed that Aβ₄₂ partially mediated the associations of APOE ɛ4 with t-Tau and p-Tau in women, with the mediation proportions of 36.13% and 42.53%, separately (Fig. 4B). However, Aβ₄₂ fully mediated the associations of APOE ɛ4 with t-Tau and p-Tau in men, with the mediation proportions of 62.85% and 70.97%, separately (Fig. 4C). In the further age-stratified mediation analyses, Aβ₄₂ still partially mediated the associations of APOE ɛ4 with p-Tau and t-Tau, regardless of age (Fig. 4D, E).

DISCUSSION

In our study, we investigated the associations of APOE ɛ2 and APOE ɛ4 genotypes with CSF AD biomarkers in individuals without dementia. Our study suggested that APOE ɛ2 genotypes had a negative association with Aβ but no association with tau. However, the destructive effects of APOE ɛ4 genotype on AD might be associated with the formation of Aβ and tau pathology. Moreover, it was interesting to note that the association between the APOE ɛ4 genotype and tau might be partially mediated by Aβ. Furthermore, our study suggested that the significant protective effects of APOE ɛ2 genotype might be modified by sex and age, but the destructive effects of APOE ɛ4 genotype might not be affected by sex and age.

We found that APOE ɛ2 genotype was protective against AD via CSF Aβ rather than tau and clinical phenotypes in non-demented individuals, which was consistent with previous studies [37, 38]. As for tau, no association between APOE ɛ2 genotype and tau [39, 40], while others reported lower levels of tau in APOE ɛ2 carriers [17]. Differences in methodology across studies could likely explained this disparity. For example, some previous studies only compared APOE ɛ2 carriers with participants without APOE ɛ2 (including APOE ɛ4 carriers), and some included very low numbers of APOE ɛ2 carriers (n < 5) [7, 41]. Moreover, we also investigated the associations of APOE ɛ2 genotype with hippocampal volumes and cognition, but we failed to find any statistically significant association, which was consistent with a previous multimodal study [37]. Besides, we tried to expand sample size (based on CSF biomarkers) and included the new ADNI composite measures, which have been shown to predict diagnostic transition from MCI to AD dementia and correlates with MRI and CSF measurements [36].

Contrary to APOE ɛ2 genotype, APOE ɛ4 carriers showed lower Aβ levels, but higher p-Tau and t-Tau levels. Results related to Aβ were in accordance with previous studies [2 , 5]. Besides, multiple neuropathological studies have also shown higher neurofibrillary tangle pathology in APOE ɛ4 carriers [7, 19]. However, studies on the effects of APOE ɛ4 genotype on CSF tau were rather scarce, with inconsistent results. These disparities might be associated with disease stage of participants, with APOE ɛ4 carriers having higher tau burden in early clinical stages but not in more advanced stage [42, 43]. Taken together, our study was largely in line with established neuropathological findings and it provided more evidence to support the association of the APOE ɛ4 allele with tau pathology. Regarding clinical phenotypes, APOE ɛ4 genotype could cause the hippocampus atrophy, hypo-metabolism, and cognitive decline. This was in line with a recent meta-analysis showing that APOE ɛ4 carriers had worse cognition than non-carriers in the memory domain [44]. However, a recent PET analysis indicated that the APOE ɛ4 genotype were not associated with any cognitive domain, which might be confounded by its small sample size. To our knowledge, some previous neuropathological and PET studies have raised the question whether APOE ɛ4 impacts tau pathology directly or through the Aβ pathway [17, 18]. Our results suggested that Aβ levels significantly mediated the association between APOE ɛ4 genotypes and tau pathology. This reminded us that the association between APOE ɛ4 genotypes and tau burden might mainly be driven by increased Aβ pathology. When it came to the association between APOE ɛ2/ɛ4 genotypes and AD, whether confounding effects existed remained controversial because the simultaneous carriage of both the ɛ2 and ɛ4 alleles is rare—approximately 2% of the population [45]. With the expansion of sample size, a recent large-scale Aβ PET imaging study found that the Aβ deposition in APOE ɛ2/ɛ4 participants were intermediate between that of ɛ2 and ɛ4 carriers, suggesting that the ɛ2 allele might counteract some of the detrimental effects of the ɛ4 allele [46]. Due to the low sample size in our study (n = 13), we failed to investigate the association between the APOE ɛ2/ɛ4 genotype and CSF biomarkers, which remained to be determined in larger samples.

Previous interaction analyses of APOE genotypes with AD phenotypes (especially CSF biomarkers and neuropathological findings) yielded contradictory result. As for the effects of sex on the associations of APOE genotypes and AD biomarkers, our study only found a positive association of APOE ɛ2 genotype with Aβ in females. These preliminary results might have some implications for further treatment. Due to our small sample size of APOE ɛ2 carriers, our results still need replication in additional large cohort studies. As for APOE ɛ4 carriers, we found that APOE ɛ4 genotypes could increase tau levels more in women than in men which was consistent with a previous study [26]. On the contrary, two neuropathological studies did not find a significant sex*APOE ɛ4 interactions on tau pathology [47, 48]. Notably, a recent large-scale CSF study showed that this interaction might only occur in the early phases of the disease (subjective cognitive decline and MCI) but not in advanced stages (AD), which might explain why a neuropathological study found no interaction [49]. As for the effects of age on the associations between APOE genotypes and AD biomarkers, our study only found a significantly positive relationship of APOE ɛ2 genotypes with Aβ in older individuals, which might have some implications for further clinical studies. The effects of APOE ɛ4 genotypes might be independent of age, similar to sex. Suggestively, in our mediation analysis stratified by sex, we found partial mediation effects in women but fully mediated effects in men. This suggested that the female-specific mediation of tau levels might partially stem from difference in sex hormones. A recent study showed that testosterone had a protective role against tau particularly among APOE ɛ4 carriers, and that low testosterone levels were more characteristic of women than men which might make a person more susceptible to tau [50]. Besides, previous animal studies showed that testosterone protected rats against the hyper-phosphorylation of tau. Thus, we could infer that the partial mediation effect in female might be caused by the typically lower testosterone levels in women [51 –53]. Certainly, further investigation is warranted to clarify this gender difference.

Besides the effect of sex and age, more and more environmental factors were gradually shown to have influence on the association between APOE and AD. Cook et al found that high educational levels were shown to interact the effects of the APOE ɛ4 genotypes on cognition in elderly individuals [54]. A meta-analysis of 37 studies showed that increased risk of AD caused by smoking was more pronounced in APOE ɛ4 non-carriers than in carriers [55]. Additionally, other factors such as cardiovascular health and healthy lifestyles were revealed to be associated with dementia regardless of APOE. To sum up, the effect of environmental factors on the association between APOE and AD still need to be explored more specifically and extensively.

The main strengths of this study were as follows. Firstly, we further conducted subgroup analyses stratified by sex and age to deeply explore the associations of APOE genotypes and AD biomarkers. Secondly, the finding of mediation analysis indicated that the positive association between APOE ɛ4 carriers and t-Tau/p-Tau might be largely explained by the effect of APOE ɛ4 carriers on Aβ. There were also several limitations. Firstly, we could not conduct a gene dosage analysis because of the limited number of APOE ɛ2 homozygotes (n = 1). Secondly, we failed to investigate whether APOE ɛ2/ɛ4 genotype played an intermediate role due to the small sample size of APOE ɛ2/ɛ4 carriers (n = 13). The effects of this gene combination on CSF biomarkers remained to be determined in larger samples. Lastly, we did not find an association of APOE ɛ2 genotype with tau pathology, so we could not conduct a mediation analysis to explore whether Aβ mediated the association between APOE ɛ2 genotype and tau burden.

Conclusion

APOE ɛ2 genotypes were associated with higher Aβ but not tau, suggesting that APOE ɛ2 genotypes could promote the clearance of Aβ. However, the destructive effects of APOE ɛ4 genotypes were correlated with the accumulation of both Aβ and tau pathology. Suggestively, the possible modification effects of sex and age on APOE ɛ2 but APOE ɛ4 carriers might have some implications for precision prevention and AD-modifying therapeutic trials.

FUNDING

This study was supported by grants from the National Natural Science Foundation of China (81971032, 81801274, and 81901121).

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The dataset generated and analyzed in the current study is available from the corresponding author on reasonable request.

Footnotes

ACKNOWLEDGMENTS

Data collection and sharing for ADNI data section was funded by the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (National Institutes of Health Grant U01 AG024904) and DOD ADNI (Department of Defense award number W81XWH-12-2-0012). ADNI is funded by the National Institute on Aging, the National Institute of Biomedical Imaging and Bioengineering, and through generous contributions from the following: AbbVie, Alzheimer’s Association; Alzheimer’s Drug Discovery Foundation; Araclon Biotech; BioClinica, Inc.; Biogen; Bristol-Myers Squibb Company; CereSpir, Inc.; Cogstate; Eisai Inc.; Elan Pharmaceuticals, Inc.; Eli Lilly and Company; EuroImmun; F. Hoffmann-La Roche Ltd and its affiliated company Genentech, Inc.; Fujirebio; GE Healthcare; IXICO Ltd.; Janssen Alzheimer Immunotherapy Research & Development, LLC.; Johnson & Johnson Pharmaceutical Research & Development LLC.; Lumosity; Lundbeck; Merck & Co., Inc.; Meso Scale Diagnostics, LLC.; NeuroRx Research; Neurotrack Technologies; Novartis Pharmaceuticals Corporation; Pfizer Inc.; Piramal Imaging; Servier; Takeda Pharmaceutical Company; and Transition Therapeutics. The Canadian Institutes of Health Research is providing funds to support ADNI clinical sites in Canada. Private sector contributions are facilitated by the Foundation for the National Institutes of Health (). The grantee organization is the Northern California Institute for Peer Review for Research and Education, and the study is coordinated by the Alzheimer’s Therapeutic Research Institute at the University of Southern California. ADNI data are disseminated by the Laboratory for Neuro Imaging at the University of Southern California.

The supplementary material is available in the electronic version of this article: .

References

Belloy

, Napolioni

, Greicius

(2019) A quarter century of APOE and Alzheimer’s disease: Progress to date and the path forward. Neuron 101, 820–838.

Jack

, Wiste

, Weigand

, Knopman

, Vemuri

, Mielke

, Lowe

, Senjem

, Gunter

, Machulda

, Gregg

, Pankratz

, Rocca

, Petersen

(2015) Age, sex, and APOE ɛ4 effects on memory, brain structure, and β-amyloid across the adult life span. JAMA Neurol 72, 511–519.

Jansen

, Ossenkoppele

, Knol

, Tijms

, Scheltens

, Verhey

FRJ

, Visser

, Aalten

, Aarsland

, Alcolea

, Alexander

, Almdahl

, Arnold

, Baldeiras

, Barthel

, van Berckel

BNM

, Bibeau

, Blennow

, Brooks

, van Buchem

, Camus

, Cavedo

, Chen

, Chetelat

, Cohen

, Drzezga

, Engelborghs

, Fagan

, Fladby

, Fleisher

, van der Flier

, Ford

, Förster

, Fortea

, Foskett

, Frederiksen

, Freund-Levi

, Frisoni

, Froelich

, Gabryelewicz

, Gill

, Gkatzima

, Gómez-Tortosa

, Gordon

, Grimmer

, Hampel

, Hausner

, Hellwig

, Herukka

S-K

, Hildebrandt

, Ishihara

, Ivanoiu

, Jagust

, Johannsen

, Kandimalla

, Kapaki

, Klimkowicz-Mrowiec

, Klunk

, Köhler

, Koglin

, Kornhuber

, Kramberger

, Van Laere

, Landau

, Lee

, de Leon

, Lisetti

, Lleó

, Madsen

, Maier

, Marcusson

, Mattsson

, de Mendonça

, Meulenbroek

, Meyer

, Mintun

, Mok

, Molinuevo

, Møllergård

, Morris

, Mroczko

, Van der Mussele

, Na

, Newberg

, Nordberg

, Nordlund

, Novak

, Paraskevas

, Parnetti

, Perera

, Peters

, Popp

, Prabhakar

, Rabinovici

, Ramakers

IHGB

, Rami

, Resende de Oliveira

, Rinne

, Rodrigue

, Rodríguez-Rodríguez

, Roe

, Rot

, Rowe

, Rüther

, Sabri

, Sanchez-Juan

, Santana

, Sarazin

, Schröder

, Schütte

, Seo

, Soetewey

, Soininen

, Spiru

, Struyfs

, Teunissen

, Tsolaki

, Vandenberghe

, Verbeek

, Villemagne

, Vos

SJB

, van Waalwijk van Doorn

LJC

, Waldemar

, Wallin

ÅK

, Wiltfang

, Wolk

, Zboch

, Zetterberg

(2015) Prevalence of cerebral amyloid pathology in persons without dementia: A meta-analysis. JAMA 313, 1924–1938.

Liu

C-C

, Liu

C-C

, Kanekiyo

, Xu

, Bu

(2013) Apolipoprotein E and Alzheimer disease: Risk, mechanisms and therapy. Nat Rev Neurol 9, 106–118.

Reiman

, Chen

, Liu

, Bandy

, Yu

, Lee

, Ayutyanont

, Keppler

, Reeder

, Langbaum

JBS

, Alexander

, Klunk

, Mathis

, Price

, Aizenstein

, DeKosky

, Caselli

(2009) Fibrillar amyloid-beta burden in cognitively normal people at 3 levels of genetic risk for Alzheimer’s disease. Proc Natl Acad Sci U S A 106, 6820–6825.

Buckley

, Mormino

, Chhatwal

, Schultz

, Rabin

, Rentz

, Acar

, Properzi

, Dumurgier

, Jacobs

, Gomez-Isla

, Johnson

, Sperling

, Hanseeuw

(2019) Associations between baseline amyloid, sex, and APOE on subsequent tau accumulation in cerebrospinal fluid. Neurobiol Aging 78, 178–185.

Nagy

, Esiri

, Jobst

, Johnston

, Litchfield

, Sim

, Smith

(1995) Influence of the apolipoprotein E genotype on amyloid deposition and neurofibrillary tangle formation in Alzheimer’s disease. Neuroscience 69, 757–761.

Oyama

, Shimada

, Oyama

, Ihara

(1995) Apolipoprotein E genotype, Alzheimer’s pathologies and related gene expression in the aged population. Brain Res Mol Brain Res 29, 92–98.

Slot

RER

, Verfaillie

SCJ

, Overbeek

, Timmers

, Wesselman

LMP

, Teunissen

, Dols

, Bouwman

, Prins

, Barkhof

, Lammertsma

, Van Berckel

BNM

, Scheltens

, Sikkes

SAM

, Van der Flier

(2018) Subjective Cognitive Impairment Cohort (SCIENCe): Study design and first results. Alzheimers Res Ther 10, 76.

10.

Therriault

, Benedet

, Pascoal

, Mathotaarachchi

, Chamoun

, Savard

, Thomas

, Kang

, Lussier

, Tissot

, Parsons

, Qureshi

MNI

, Vitali

, Massarweh

, Soucy

J-P

, Rej

, Saha-Chaudhuri

, Gauthier

, Rosa-Neto

(2020) Association of Apolipoprotein E ɛ4 with medial temporal tau independent of amyloid-β. JAMA Neurol 77, 470–479.

11.

Yamazaki

, Zhao

, Caulfield

, Liu

C-C

, Bu

(2019) Apolipoprotein E and Alzheimer disease: Pathobiology and targeting strategies. Nat Rev Neurol 15, 501–518.

12.

Reiman

, Arboleda-Velasquez

, Quiroz

, Huentelman

, Beach

, Caselli

, Chen

, Su

, Myers

, Hardy

, Paul Vonsattel

, Younkin

, Bennett

, De Jager

, Larson

, Crane

, Keene

, Kamboh

, Kofler

, Duque

, Gilbert

, Gwirtsman

, Buxbaum

, Dickson

, Frosch

, Ghetti

, Lunetta

, Wang

L-S

, Hyman

, Kukull

, Foroud

, Haines

, Mayeux

, Pericak-Vance

, Schneider

, Trojanowski

, Farrer

, Schellenberg

, Beecham

, Montine

, Jun

(2020) Exceptionally low likelihood of Alzheimer’s dementia in APOE2 homozygotes from a 5,000-person neuropathological study. Nat Commun 11, 667.

13.

Suri

, Heise

, Trachtenberg

, Mackay

(2013) The forgotten APOE allele: A review of the evidence and suggested mechanisms for the protective effect of APOE ɛ2. Neurosci Biobehav Rev 37, 2878–2886.

14.

Corder

, Saunders

, Risch

, Strittmatter

, Schmechel

, Gaskell

PCJ

, Rimmler

, Locke

, Conneally

, Schmader

(1994) Protective effect of apolipoprotein E type 2 allele for late onset Alzheimer disease. Nat Genet 7, 180–184.

15.

Chiang

, Insel

, Tosun

, Schuff

, Truran-Sacrey

, Raptentsetsang

, Jack

, Aisen

, Petersen

, Weiner

(2010) Hippocampal atrophy rates and CSF biomarkers in elderly APOE2 normal subjects. Neurology 75, 1976–1981.

16.

Conejero-Goldberg

, Gomar

, Bobes-Bascaran

, Hyde

, Kleinman

, Herman

, Chen

, Davies

, Goldberg

(2014) APOE2 enhances neuroprotection against alzheimer’s disease through multiple molecular mechanisms. Mol Psychiatry 19, 1243–1250.

17.

Serrano-Pozo

, Qian

, Monsell

, Betensky

, Hyman

(2015) APOEɛ2 is associated with milder clinical and pathological Alzheimer disease. Ann Neurol 77, 917–929.

18.

Mungas

, Tractenberg

, Schneider

, Crane

, Bennett

(2014) A 2-process model for neuropathology of Alzheimer’s disease. Neurobiol Aging 35, 301–308.

19.

Farfel

, Yu

, De Jager

, Schneider

, Bennett

(2016) Association of APOE with tau-tangle pathology with and without β-amyloid. Neurobiol Aging 37, 19–25.

20.

Shi

, Yamada

, Liddelow

, Smith

, Zhao

, Luo

, Tsai

, Spina

, Grinberg

, Rojas

, Gallardo

, Wang

, Roh

, Robinson

, Finn

, Jiang

, Sullivan

, Baufeld

, Wood

, Sutphen

, McCue

, Xiong

, Del-Aguila

, Morris

, Cruchaga

, Fagan

, Miller

, Boxer

, Seeley

, Butovsky

, Barres

, Paul

, Holtzman

(2017) ApoE4 markedly exacerbates tau-mediated neurodegeneration in a mouse model of tauopathy. Nature 549, 523–527.

21.

van der Kant

, Goldstein

LSB

, Ossenkoppele

(2020) Amyloid-β-independent regulators of tau pathology in Alzheimer disease. Nat Rev Neurosci 21, 21–35.

22.

Andersen

, Launer

, Dewey

, Letenneur

, Ott

, Copeland

, Dartigues

, Kragh-Sorensen

, Baldereschi

, Brayne

, Lobo

, Martinez-Lage

, Stijnen

, Hofman

(1999) Gender differences in the incidence of AD and vascular dementia: The EURODEM Studies. EURODEM Incidence Research Group. Neurology 53, 1992–1997.

23.

Barnes

, Wilson

, Bienias

, Schneider

, Evans

, Bennett

(2005) Sex differences in the clinical manifestations of Alzheimer disease pathology. Arch Gen Psychiatry 62, 685–691.

24.

Koran

MEI

, Wagener

, Hohman

(2017) Sex differences in the association between AD biomarkers and cognitive decline. Brain Imaging Behav 11, 205–213.

25.

Neu

, Pa

, Kukull

, Beekly

, Kuzma

, Gangadharan

, Wang

L-S

, Romero

, Arneric

, Redolfi

, Orlandi

, Frisoni

, Au

, Devine

, Auerbach

, Espinosa

, Boada

, Ruiz

, Johnson

, Koscik

, Wang

J-J

, Hsu

W-C

, Chen

Y-L

, Toga

(2017) Apolipoprotein E genotype and sex risk factors for Alzheimer disease: A meta-analysis. JAMA Neurol 74, 1178–1189.

26.

Altmann

, Tian

, Henderson

, Greicius

(2014) Sex modifies the APOE-related risk of developing Alzheimer disease. Ann Neurol 75, 563–573.

27.

Saddiki

, Fayosse

, Cognat

, Sabia

, Engelborghs

, Wallon

, Alexopoulos

, Blennow

, Zetterberg

, Parnetti

, Zerr

, Hermann

, Gabelle

, Boada

, Orellana

, de Rojas

, Lilamand

, Bjerke

, van Broeckhoven

, Farotti

, Salvadori

, Diehl-Schmid

, Grimmer

, Hourregue

, Dugravot

, Nicolas

, Laplanche

, Lehmann

, Bouaziz-Amar

, Hugon

, Tzourio

, Singh-Manoux

, Paquet

, Dumurgier

(2020) Age and the association between apolipoprotein E genotype and Alzheimer disease: A cerebrospinal fluid biomarker–based case–control study. PLoS Med 17, e1003289.

28.

Folstein

, Folstein

, McHugh

(1975) “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

29.

Morris

(1993) The Clinical Dementia Rating (CDR): Current version and scoring rules. Neurology 43, 2412–2414.

30.

Y-N

, Xu

, Li

J-Q

, Guo

, Cui

, Chen

K-L

, Huang

Y-Y

, Dong

, Tan

, Yu

J-T

(2019) FDG-PET as an independent biomarker for Alzheimer’s biological diagnosis: A longitudinal study. Alzheimers Res Ther 11, 57.

31.

Saykin

, Shen

, Foroud

, Potkin

, Swaminathan

, Kim

, Risacher

, Nho

, Huentelman

, Craig

, Thompson

, Stein

, Moore

, Farrer

, Green

, Bertram

, Jack

CRJ

, Weiner

(2010) Alzheimer’s Disease Neuroimaging Initiative biomarkers as quantitative phenotypes: Genetics core aims, progress, and plans. Alzheimers Dement 6, 265–273.

32.

Gosche

, Mortimer

, Smith

, Markesbery

, Snowdon

(2002) Hippocampal volume as an index of Alzheimer neuropathology: Findings from the Nun Study. Neurology 58, 1476–1482.

33.

Jack

CRJ

, Petersen

, Xu

, O’Brien

, Smith

, Ivnik

, Boeve

, Waring

, Tangalos

, Kokmen

(1999) Prediction of AD with MRI-based hippocampal volume in mild cognitive impairment. Neurology 52, 1397–1403.

34.

Braak

, Braak

(1998) Evolution of neuronal changes in the course of Alzheimer’s disease. J Neural Transm Suppl 53, 127–140.

35.

Crane

, Carle

, Gibbons

, Insel

, Mackin

, Gross

, Jones

, Mukherjee

, Curtis

, Harvey

, Weiner

, Mungas

(2012) Development and assessment of a composite score for memory in the Alzheimer’s Disease Neuroimaging Initiative (ADNI). Brain Imaging Behav 6, 502–516.

36.

Choi

S-E

, Mukherjee

, Gibbons

, Sanders

, Jones

, Tommet

, Mez

, Trittschuh

, Saykin

, Lamar

, Rabin

, Foldi

, Sikkes

, Jutten

, Grandoit

, Mac

Donald C

, Risacher

, Groot

, Ossenkoppele

, Crane

(2020) Development and validation of language and visuospatial composite scores in ADNI. Alzheimers Dement (N Y) 6, e12072.

37.

Grothe

, Villeneuve

, Dyrba

, Bartrés-Faz

, Wirth

(2017) Multimodal characterization of older APOE2carriers reveals selective reduction of amyloid load. Neurology 88, 569–576.

38.

Toledo

, Zetterberg

, van Harten

, Glodzik

, Martinez-Lage

, Bocchio-Chiavetto

, Rami

, Hansson

, Sperling

, Engelborghs

, Osorio

, Vanderstichele

, Vandijck

, Hampel

, Teipl

, Moghekar

, Albert

, Hu

, Monge Argilés

, Gorostidi

, Teunissen

, De Deyn

, Hyman

, Molinuevo

, Frisoni

, Linazasoro

, de Leon

, van der Flier

, Scheltens

, Blennow

, Shaw

, Trojanowski

(2015) Alzheimer’s disease cerebrospinal fluid biomarker in cognitively normal subjects. Brain 138, 2701–2715.

39.

Lippa

, Smith

, Saunders

, Hulette

, Pulaski-Salo

, Roses

(1997) Apolipoprotein E-epsilon 2 and Alzheimer’s disease: Genotype influences pathologic phenotype. Neurology 48, 515–519.

40.

Tiraboschi

, Hansen

, Masliah

, Alford

, Thal

, Corey-Bloom

(2004) Impact of APOE genotype on neuropathologic and neurochemical markers of Alzheimer disease. Neurology 62, 1977–1983.

41.

Morris

, Benjamin

, Leake

, McArthur

, Candy

, Ince

, Torvik

, Bjertness

, Edwardson

(1995) Effect of apolipoprotein E genotype on Alzheimer’s disease neuropathology in a cohort of elderly Norwegians. Neurosci Lett 201, 45–47.

42.

Ossenkoppele

, van der Flier

, Zwan

, Adriaanse

, Boellaard

, Windhorst

, Barkhof

, Lammertsma

, Scheltens

, van Berckel

BNM

(2013) Differential effect of APOE genotype on amyloid load and glucose metabolismin AD dementia. Neurology 80, 359–365.

43.

Lehmann

, Ghosh

, Madison

, Karydas

, Coppola

, O’Neil

, Huang

, Miller

, Jagust

, Rabinovici

(2014) Greater medial temporal hypometabolism and lower cortical amyloid burden in ApoE4-positive AD patients. J Neurol Neurosurg Psychiatry 85, 266–273.

44.

Emrani

, Arain

, DeMarshall

, Nuriel

(2020) APOE4 is associated with cognitive and pathological heterogeneity in patients with Alzheimer’s disease: A systematic review. Alzheimers Res Ther 12, 141.

45.

Mahley

(1988) Apolipoprotein E: Cholesterol transport protein with expanding role in cell biology. Science 240, 622–630.

46.

Insel

, Hansson

, Mattsson-Carlgren

(2021) Association between apolipoprotein E ɛ2 vs ɛ4, age, and β-amyloid in adults without cognitive impairment. JAMA Neurol 78, 229–235.

47.

Oveisgharan

, Buchman

, Yu

, Farfel

, Hachinski

, Gaiteri

, De Jager

, Schneider

, Bennett

(2018) APOE ɛ2ɛ4 genotype, incident AD and MCI, cognitive decline, and AD pathology in older adults. Neurology 90, e2127–e2134.

48.

Hohman

, Dumitrescu

, Barnes

, Thambisetty

, Beecham

, Kunkle

, Gifford

, Bush

, Chibnik

, Mukherjee

, et al. (2018) Sex-specific association of apolipoprotein e with cerebrospinal fluid levels of tau. JAMA Neurol 75, 989–998.

49.

Babapour Mofrad

, Tijms

, Scheltens

, Barkhof

, van der Flier

, Sikkes

SAM

, Teunissen

(2020) Sex differences in CSF biomarkers vary by Alzheimer disease stage and APOE ɛ4 genotype. Neurology 95, e2378–e2388.

50.

Sundermann

, Panizzon

, Chen

, Andrews

, Galasko

, Banks

(2020) Sex differences in Alzheimer’s-related Tau biomarkers and a mediating effect of testosterone. Biol Sex Differ 11, 33.

51.

Rosario

, Carroll

, Pike

(2010) Testosterone regulation of Alzheimer-like neuropathology in male 3xTg-AD mice involves both estrogen and androgen pathways. Brain Res 1359, 281–290.

52.

Papasozomenos

(1997) The heat shock-induced hyperphosphorylation of tau is estrogen-independent and prevented by androgens: Implications for Alzheimer disease. Proc Natl Acad Sci U S A 94, 6612–6617.

53.

Papasozomenos

, Shanavas

(2002) Testosterone prevents the heat shock-induced overactivation of glycogen synthase kinase-3 beta but not of cyclin-dependent kinase 5 and c-Jun NH2-terminal kinase and concomitantly abolishes hyperphosphorylation of tau: Implications for Alzheimer’s disease. Proc Natl Acad Sci U S A 99, 1140–1145.

54.

Cook

, Fletcher

(2015) Can education rescue genetic liability for cognitive decline? Soc Sci Med 127, 159–170.

55.

Zhong

, Wang

, Zhang

, Guo

, Zhao

(2015) Smoking is associated with an increased risk of dementia: A meta-analysis of prospective cohort studies with investigation of potential effect modifiers. PLoS One 10, e0118333.

56.

Frank

, Hensel

, Baak

, Schramm

, Dragano

, Weimar

, Hoffmann

, Nöthen

, Erbel

, Jöckel

K-H

, Jokisch

, Schmidt

(2021) Interaction of Alzheimer’s disease-associated genetic risk with indicators of socioeconomic position on mild cognitive impairment in the Heinz Nixdorf Recall Study. J Alzheimers Dis 82, 1715–1725.

57.

Jin

, He

, Zhang

, Gong

, Niu

, Ji

, Li

, Zeng

, Yan

(2021) Association of APOE ɛ4 genotype and lifestyle with cognitive function among Chinese adults aged 80 years and older: A cross-sectional study. PLoS Med 18, e1003597.

58.

Lee

, Hughes

, George

, Griswold

, Sedaghat

, Simino

, Lutsey

(2022) Education and cardiovascular health as effect modifiers of APOE ɛ4 on dementia: The Atherosclerosis Risk in Communities Study. J Gerontol A Biol Sci Med Sci 77, 1199–1207.