Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2022 Dec 20;30(3):597–605. doi: 10.1111/ene.15657

The APOE4 effect: structural brain differences in Alzheimer's disease according to the age at symptom onset

Gonzalo Forno 1,2, Jose Contador 1, Agnès Pérez‐Millan 1,3, Nuria Guillen 1, Neus Falgàs 1,4, Jordi Sarto 1, Adrià Tort‐Merino 1, Magdalena Castellví 1, Beatriz Bosch 1, Guadalupe Fernández‐Villullas 1, Mircea Balasa 1,4,5, Anna Antonell 1, Roser Sala‐ Llonch 1,3,6, Raquel Sanchez‐Valle 1, Michael Hornberger 7, Albert Lladó 1,✉
PMCID: PMC10108138  PMID: 36463489

Abstract

Background and purpose

How the APOE genotype can differentially affect cortical and subcortical memory structures in biomarker‐confirmed early‐onset (EOAD) and late‐onset (LOAD) Alzheimer's disease (AD) was assessed.

Method

Eighty‐seven cerebrospinal fluid (CSF) biomarker‐confirmed AD patients were classified according to their APOE genotype and age at onset. 28 were EOAD APOE4 carriers (+EOAD), 21 EOAD APOE4 non‐carriers (–EOAD), 23 LOAD APOE4 carriers (+LOAD) and 15 LOAD APOE4 non‐carriers (–LOAD). Grey matter (GM) volume differences were analyzed using voxel‐based morphometry in Papez circuit regions. Multiple regression analyses were performed to determine the relation between GM volume loss and cognition.

Results

Significantly more mammillary body atrophy in +EOAD compared to –EOAD is reported. The medial temporal and posterior cingulate cortex showed less GM in +LOAD compared to –LOAD. Medial temporal GM volume loss was also found in +EOAD compared to –LOAD. With an exception for +EOAD, medial temporal GM was strongly associated with episodic memory in the three groups, whilst posterior cingulate cortex GM volume was more related with visuospatial abilities. Visuospatial abilities and episodic memory were also associated with the anterior thalamic nucleus in –LOAD.

Conclusions

Our results show that the APOE genotype has a significant effect on GM integrity as a function of age of disease onset. Specifically, whilst LOAD APOE4 genotype is mostly associated with increased medial temporal and parietal atrophy compared to –LOAD, for EOAD APOE4 might have a more specific effect on subcortical (mammillary body) structures. The findings suggest that APOE genotype needs to be taken into account when classifying patients by age at onset.

Keywords: age at symptom onset, APOE4, early onset Alzheimer's disease, late onset Alzheimer's disease, Papez circuit

INTRODUCTION

There is an ongoing debate about the brain structural [1, 2, 3], cognitive [4] and neuropsychiatric [5] differences between early‐onset Alzheimer's disease (EOAD) and late‐onset Alzheimer's disease (LOAD). Evidence suggests that EOAD (age at symptom onset [AAO] <65) presents with greater temporoparietal or frontoparietal atrophy associated with worse non‐memory cognitive deficits [2, 4, 6, 7] and a more rapid decline [2], whereas LOAD (AAO ≥65) is more commonly associated with increased medial temporal atrophy and episodic memory deficits [1, 3, 6].

However, most previous studies on EOAD and LOAD did not have biomarker (amyloid, tau) confirmation in their cases, raising the possibility that conditions mimicking AD and particularly EOAD, might have inadvertently been part of those studies. Further, the APOE genotype has not been taken into account in previous studies even though it is the major genetic risk factor for sporadic AD [8] and it has been associated with reduced AAO [9]. Carrying at least one ε4 allele is associated with greater and more specific medial temporal atrophy [10] and reduced metabolism [11] compared with non‐carriers. By contrast, the absence of an ε4 allele and an AAO before 65 increases the vulnerability for whole‐brain volume loss [12]. This heterogeneity according to APOE genotype has been suggested to account for some of the brain structural differences between EOAD and LOAD. To date, studies have only analyzed how APOE4 impacts brain atrophy according to age [10, 13, 14] but not as a function of AAO.

The current study tries to add knowledge to these shortcomings by analyzing brain grey matter (GM) volume in biomarker‐confirmed EOAD and LOAD. Importantly, patients were divided according to their APOE genotype to elucidate whether carrying an APOE gene risk might contribute to the presentation of EOAD versus LOAD structural brain differences. Specifically, Papez circuit structures (hippocampus, mammillary bodies [MB], anterior thalamic nucleus [ATN], posterior cingulate cortex [PCC] and anterior parahippocampal cortex [aPHC]), which are known to be the critical network for AD pathophysiology [15], were analyzed using voxel‐based morphometry (VBM). The relationship between GM morphology and cognition was also studied with a particular focus on episodic memory and spatial orientation. It was hypothesized that APOE4 genotype contributes to the presentation of EOAD, since carrying the risk gene affects the presentation of symptom onset.

METHODS

Participants

A retrospective cohort of 93 participants was selected from the Alzheimer's Disease and Other Cognitive Disorders Unit at the Hospital Clinic de Barcelona (HCB), Barcelona, Spain. 50 participants were classified as EOAD and 43 as LOAD. All participants underwent a complete neurological and neuropsychological evaluation, 3 T brain magnetic resonance imaging (MRI) and a spinal tap for AD cerebrospinal fluid (CSF) biomarkers. Participants fulfilled the National Institute on Aging–Alzheimer's Association (NIA‐AA) diagnostic criteria for mild cognitive impairment due to AD with high likelihood, or dementia due to AD presenting the core CSF biomarker levels suggesting the presence of AD neuropathology (A+ T+) with neurodegeneration (N+) [16, 17]. Participants were further classified according to their APOE genotype profile: EOAD and LOAD carriers, with at least one ε4 allele (+EOAD and +LOAD respectively), and EOAD and LOAD non‐carriers, without an ε4 allele (–EOAD and –LOAD respectively). Of the 93 patients, one EOAD and five LOAD were excluded from further analysis due to lack of information on their APOE genotype. Also, patients with prior mental illness, head trauma, cerebrovascular disease, or alcohol and drug abuse were excluded. The final sample therefore comprised 28 +EOAD, 21 –EOAD, 23 +LOAD and 15 –LOAD (total N = 87). AAO was registered as the first cognitive complaint either self‐reported or reported by a knowledgeable informant. Disease duration was then calculated as the time between AAO and the MRI date.

This study was approved by the HCB Ethics Committee (HCB/2019/0054) and informed consent was obtained from all participants.

Cerebrospinal fluid biomarkers and APOE genotype

All participants underwent a lumbar puncture to measure AD CSF biomarkers. Levels of amyloid β (Aβ42), total tau (t‐tau) and phosphorylated tau (p‐tau) were determined using commercially available single‐analyte enzyme‐linked immunosorbent assay (ELISA) (INNOTEST; FUJIREBIO Europe N.V). Cut‐off values for CSF Aβ42 (A), p‐tau (T) and t‐tau (N) were determined following internal cut‐offs [18]. The APOE genotype was determined through rs429358 and rs7412 analyses by Sanger sequencing.

Imaging acquisition

Magnetic resonance imaging acquisition was performed in a 3 T Magneton Trio Tim scanner (Siemens Medical Systems) at the Magnetic Resonance Imaging Core Facility. Whole‐brain high‐resolution T1‐weighted anatomical 3D spin echo volumes, parallel to the plane connecting the anterior and posterior commissures, were used to generate 240 contiguous slices with the following parameters: repetition time 2300 ms; echo time 2.98 ms; field‐of‐view 256 mm; matrix dimension 256 × 256; slice thickness 1 mm, voxel size 1 × 1 × 1 mm3.

Imaging processing and voxel‐based morphometry analysis

Magnetic resonance imaging data were preprocessed using the DARTEL Toolbox [19] via Statistical Parametric Mapping software (SPM12) https://www.fil.ion.ucl.ac.uk/spm/software/spm12/. First, T1‐weighted images in native space were segmented into GM, white matter (WM) and CSF. Segmented GM and WM were used to run the “DARTEL (create template)” using the default parameters indicated by SPM12. Next, the “Normalized to MNI space” module from the DARTEL tools was selected to affine register the last template from the previous step into MNI space. All images were modulated using Jacobian determinants to avoid bias in the intensity of an area due to image expansion during warping and to correct volume changes. Finally, an isotropic Gaussian kernel with a standard deviation of 3 mm (full‐width at half‐maximum 8 mm) was applied to all modulated images. Regions of interest (ROIs) for the hippocampus, MB, ATN, PCC and the aPHC were created using an automated anatomical labeling (AAL3) tool [20], the TD Brodmann areas+ and the SPM brain template in the WFU_PickAtlas toolkit https://www.nitrc.org/projects/wfu_pickatlas/. ROIs were then combined for the creation of the Papez circuit by applying a binary mask. After spatial preprocessing, the smoothed, modulated, normalized GM datasets were used for statistical analysis. Images were analyzed within general linear models for second level analyses http://www.fil.ion.ucl.ac.uk/spm/software/spm12.

Further, a two‐sample t test between pairs of groups was performed to account for global atrophy across patients, correcting by total intracranial volume and age when they correspond (i.e., ε4EOAD vs. ε4LOAD). The analysis was then repeated for each ROI following the same objective, given a total of six contrasts per analysis. Contrasts are presented either with a threshold of p < 0.05 using correction for multiple comparisons with family‐wise error (FWE) correction space [21] or with an uncorrected threshold of p < 0.001 with a cluster threshold of at least 50 contiguous voxels [22, 23].

Finally, multiple regression analyses were performed using a general linear model within each group to determine the relationship between morphometry and cognitive performance (memory and visuospatial abilities). This analysis was conducted with every Papez substructure using the Free and Cued Selective Reminding Test (FCSRT) [24] and the Visual Object and Space Perception Battery (VOSP) [25] as variables of interest. Results were considered significant if they survived an FWE‐corrected threshold of p < 0.05 or a spatial threshold of 50 voxels with an uncorrected statistical threshold of p < 0.001 [23].

Statistical analysis

The Shapiro–Wilk test was used to check for normality distribution. Normally distributed data were compared across the four groups via one‐way ANOVA followed by Tukey post hoc tests. Non‐normally distributed scores were compared using Kruskal–Wallis tests. When significant differences were found, a Dunn test was used to compare specific groups. All statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) version 25 for Windows [26].

RESULTS

Demographics and cognition

Demographic and neuropsychological data are summarized in Table 1. No significant differences were found in AAO and disease duration on comparing +EOAD versus –EOAD and +LOAD versus –LOAD. After multiple comparison correction, significant differences were found in memory function, with +EOAD performing worse than –LOAD and +LOAD performing worse than –LOAD; EOAD, both carriers and non‐carriers, performed worse than +LOAD in visuospatial abilities. No significant differences were found in education, Mini‐Mental State Examination, attention, executive functions, verbal fluency, naming and comprehension abilities. Aβ42 was the only CSF biomarker that differed significantly between +LOAD and –EOAD, with LOAD carriers showing less Aβ42 compared with EOAD non‐carriers. No other CSF biomarker appears to significantly differentiate across groups.

TABLE 1.

Demographic and cognitive test results

–EOAD (n = 21) Mean (SD) +EOAD (n = 28) Mean (SD) –LOAD (n = 15) Mean (SD) +LOAD (n = 23) Mean (SD) Corrected p value
Age at MRI † 58.57 (4.52)c,d 61.18 (3.8)c,d 74.6 (5.18)a,b 74.91 (4.47)a,b 0.001a,b,c,d
Age at symptom onset 55.61 (4.35)c,d 57.54 (3.65)c,d 71 (4.24)a,b 72.23 (4.47)a,b 0.001a,b,c,d
Disease duration 2.95 (1.66) 4.07 (2.42) 2.82 (2.52) 2.89 (2.22) 0.175
Education 10.10 (4.17) 11.11 (5.15) 9.07 (4.57) 11.10 (3.65) 0.380
Sex (f:m) 13:8 14:14 12:3 7:16 0.020
MMSE † 22.20 (3.82) 22.15 (3.66) 23.93 (4.34) 24.05 (2.82) 0.187
FCSRT_FL † 9.50 (7.12) 5.95 (5.1)c 14.09 (9.31)b,d 7.14 (5.91)c 0.009b,c; 0.033c,d
FCSRT_TL 24.78 (13.17) 13.20 (9.27) 27.40 (16.7) 19.76 (13.34) 0.062
FCSRT_DFR 3.22 (3.86) 1.60 (2.82) 4.10 (4.43) 1.38 (2.56) 0.069
FCSRT_DTR 7.89 (5.81) 3.65 (4.20) 8.10 (6.40) 5.67 (4.90) 0.057
Semantic fluency 13 (4.75) 11.18 (3.83) 14.86 (8.6) 13.77 (3.68) 0.105
FAS 24.27 (15.36) 21.38 (10.82) 21.77 (15.15) 26.95 (11.19) 0.389
TMT_A 83.06 (42.92) 101.81 (51.8) 96.36 (52.65) 78.95 (38.81) 0.435
TMT_B 218 (88.73) 242.75 (72.75) 236.5 (75.69) 200.5 (76.75) 0.511
BNT 47.16 (11.64) 44.55 (10.86) 38.36 (16.51) 48.82 (7.33) 0.192
BDAE_Comprehension 13.89 (2.22) 14.5 (0.67) 14 (1.57) 14.73 (0.55) 0.511
VOSP_Letters 15.44 (6.77) 17.38 (4.57) 18.14 (2.07) 19.36 (1.05) 0.114
VOSP_Numbers 6.24 (3.62)d 6.95 (3.09)d 8.08 (2.84) 9.14 (0.89)a,b 0.008a,d; 0.047b,d.
CSF_Aβ42 434.79 (150.39)d 384.58 (106.04) 368.92 (84.66) 333.64 (91.57)a 0.043
CSF_Tau 906.92 (456.11) 887.18 (377.38) 942.11 (647.10) 836.05 (350.88) 0.94
CSF_pTau 126.26 (70.81) 116.05 (32.23) 121.77 (53.21) 116.68 (46.27) 0.9
Open in a new tab

Notes: Missing data for age at symptom onset: four –LOAD, six +LOAD. Missing data for disease duration: four –LOAD, six +LOAD. Missing data for education: one –EOAD, one +EOAD, one –LOAD, two +LOAD. Missing data for MMSE: one –EOAD, two +EOAD, one –LOAD, three +LOAD. Missing data for FCSRT_FL: three –EOAD, eight +EOAD, four –LOAD, two +LOAD. Missing data for FCSRT_TL: three –EOAD, eight +EOAD, five –LOAD, two +LOAD. Missing data for FCSRT_DFR: three –EOAD, eight +EOAD, five –LOAD, two +LOAD. Missing data for FCSRT_DTR: three –EOAD, eight +EOAD, five –LOAD, two +LOAD. Missing data for semantic fluency: two –EOAD, six +EOAD, one –LOAD, one +LOAD. Missing data for FAS: six –EOAD, seven +EOAD, two –LOAD, one +LOAD. Missing data for TMT_A: three –EOAD, seven +EOAD, one –LOAD, one +LOAD. Missing data for TMT_B: 12 –EOAD, 16 +EOAD, five –LOAD, seven +LOAD. Missing data for BNT: two –EOAD, six +EOAD, one –LOAD, one +LOAD. Missing data for BDAE_Comprehension: three –EOAD, six +EOAD, one –LOAD, one +LOAD. Missing data for VOSP_Letters: three –EOAD, seven +EOAD, one –LOAD, one +LOAD. Missing data for VOSP_Numbers: four –EOAD, eight +EOAD, two –LOAD, one +LOAD.

Abbreviations: BDAE_Comprehension, Boston Diagnostic Aphasia Examination comprehension; BNT, Boston Naming Test; CSF_Aβ42, cerebrospinal fluid amyloid β42; CSF_Tau, cerebrospinal fluid total tau; CSF_pTau, cerebrospinal fluid phosphorylated tau; +EOAD, early‐onset APOE4 carriers; –EOAD, early‐onset APOE4 non‐carriers; FAS, phonemic fluency; FCSRT_DFR, Free and Cued Selective Reminding Test delayed free recall; FCSRT_DTR: Free and Cued Selective Reminding Test delayed total recall; FCSRT_FL, Free and Cued Selective Reminding Test free learning; FCSRT_TL, Free and Cued Selective Reminding Test total learning; +LOAD, late‐onset APOE4 carriers; –LOAD, late‐onset APOE4 non‐carriers; MMSE, Mini‐Mental State Examination; MRI, magnetic resonance imaging; SD, standard deviation; TMT_A, trail making test A; TMT_B, trail making test B; VOSP_Letters, Visual Object and Space Perception Battery letters; VOSP_Numbers, Visual Object and Space Perception Battery numbers.

†

ANOVA test: p values refer to the Tukey post hoc.

Kruskal–Wallis: p values refer to the Dunn test.aDiffer from –EOAD, bDiffer from +EOAD, cDiffer from –LOAD, dDiffer from +LOAD.

Group differences in GM intensity

For the whole‐brain analyses and the Papez circuit mask, see Supplementary Material 1. Within the Papez circuit substructures, the following GM differences were observed (see Figure 1).

  1. +EOAD versus –EOAD: +EOAD showed MB atrophy as the sole affected region compared to –EOAD (p < 0.05, FWE corrected), with no other contrast showing similar results.

  2. +LOAD versus –LOAD: +LOAD showed significant PCC and hippocampal atrophy (both p values, uncorrected p < 0.001) compared to –LOAD with no other regions emerging to a significant level.

  3. +EOAD versus –LOAD: +EOAD showed significant hippocampal and aPHC decreased GM volume after p < 0.05, FWE corrected, compared with –LOAD.

FIGURE 1.

FIGURE 1

Open in a new tab

Atrophy—group comparisons

No other contrasts between groups showed GM differences.

GM intensity and cognition

Memory and visuospatial abilities correlation with GM decreased volume are shown in Figure 2.

  1. +EOAD showed no significant associations between cognitive impairment and any Papez circuit atrophy.

  2. –EOAD episodic memory impairment was strongly associated with medial temporal areas (FCSRT immediate free recall, FCSRT total immediate recall and total delayed recall, p < 0.05, FWE corrected). Further, visuospatial performance was associated with medial temporal atrophy (VOSP letters and numbers perception, p < 0.05, FWE corrected) and PCC atrophy (VOSP letters perception, p < 0.05, FWE corrected).

  3. +LOAD showed significant association between memory impairment and Papez circuit GM volume decrease, with several medial temporal areas correlating with episodic memory scores (FCSRT overall, all p < 0.05, FWE corrected). Further, PCC atrophy was associated with visuospatial abilities (VOSP letters perception, p < 0.05, FWE corrected).

  4. –LOAD showed significant association of medial temporal atrophy with episodic memory performance (FCSRT overall, all p < 0.05, FWE corrected), whilst FCSRT free recall and delayed free recall were associated with PCC atrophy (both p values, uncorrected p < 0.001). Episodic memory was also associated with ATN GM volume decrease (FCSRT total immediate recall, p < 0.05, FWE corrected). ATN (VOSP letters and numbers perception, p < 0.05, FWE corrected) and PCC atrophy were associated with visuospatial impairment (VOSP numbers perception, p < 0.05, FWE corrected).

FIGURE 2.

FIGURE 2

Open in a new tab

Atrophy—cognitive correlates

DISCUSSION

Our results show that the APOE genotype differentially affects Papez circuit structures as a function of EOAD and LOAD. This was further confirmed by the cognitive correlates which show differential effects across the APOE carrier. The results suggest that the APOE genotype should be taken into consideration for research and treatment studies as it might differentially affect not only the underlying brain structural difference but also associated cognitive symptomology.

In more detail, significant MB atrophy was found in +EOAD compared to –EOAD, which is somewhat surprising as MB atrophy findings in AD have been inconsistent. Whilst some of the previous studies have shown MB shrinkage in AD compared to healthy controls [27, 28], others have been unable to replicate those findings [21]. It is well established that MB is a critical part of the Papez circuit network and has an important role in episodic memory [15, 29], particularly due to its efferents from the hippocampus. Therefore, it makes sense that MB could have a role in AD pathophysiology but its small size might mean that finding could be inconsistent due to the extent of the atrophy detected; future investigation is required. As only the +EOAD group showed these changes, previous inconsistent results might be due to different admixtures of APOE4 and APOE4 non‐carriers. Still, it needs to be noted that the ε4 carriers MB effect was only present for +EOAD and not +LOAD. This clearly needs to be addressed in further investigations.

By contrast, +LOAD showed no MB differences compared to –LOAD; instead significantly more GM volume loss in medial temporal areas and PCC was predominant. These findings align with several studies indicating a clear influence of ε4 on the expression of neuropathology [7]. MRI studies have shown more severe brain atrophy in AD carriers compared with non‐carriers in medial temporal areas [10, 14]. Parietal, temporal and PCC hypometabolism has also been reported in +LOAD compared with –LOAD [30]. Our results replicate those previous findings to a large degree.

In terms of the cognitive correlates, no significant results were found in cognitive–GM correlations for +EOAD. Our study is not an exception as previous studies analyzing structural and functional changes in APOE4 carriers have shown conflicting results [10, 31]. Similarly, potential ceiling or floor effects for +EOAD could explain this null result. Experimental cognitive tests in particular targeting spatial navigation deficits have shown differences for APOE4 [32, 33], whereas even experimental memory tests often do not report differences according to APOE genotype [34]. Further, –EOAD temporoparietal atrophy was highly associated with episodic memory and visuospatial impairment—results that are not surprising due to the close relation between PCC and medial temporal regions and their role in cognition [35, 36].

For the LOAD patients, both APOE carriers and non‐carriers showed memory and visuospatial impairment associated with medial temporal and PCC atrophy. Notably, ATN atrophy also appears to be related to memory and visuospatial deficits in –LOAD. Although ATN altered functional connectivity has been related to poor performance in memory tasks in mild cognitive impairment carriers [37, 38], our results only showed decreased ATN volume and memory impairment in –LOAD but not in +LOAD. Previous studies have shown that functional impairment can occur independently of GM volume changes [37, 39], suggesting that ATN‐related memory impairment in APOE4 carriers could respond more to a functional impairment than atrophy per se. This is clearly a striking contrast to +EOAD findings, for which another limbic structure (MB) emerged as the critical difference across genotypes. By contrast, for LOAD the non‐carriers seem to show cognitive–atrophy correlates for the ATN. The exact role of how the ATN contributes to symptomatology in AD is still poorly understood, despite it being known since the earliest Braak and Braak studies that it is one of the first regions to be affected by AD pathophysiology [40]. Still, our results indicate that the ATN might have an important role in the onset of visuospatial and memory deficits in LOAD.

Clinically, our results suggest an interaction of the APOE genotype with AAO in AD, which not only is associated with differential atrophy but also results in different cognitive impairments. Whilst APOE4 carriers showed increased episodic memory impairment, APOE4 non‐carriers present greater visuospatial deficits, in particular for EOAD. This is in line with previous studies showing that carriers of the ε4 allele had greater memory impairment than non‐carriers [41] and EOAD presents with increased visuospatial deficits [6]. Further, carriers have been related with somehow specific medial temporal atrophy [10, 14], whilst non‐carriers show whole‐brain volume vulnerability [7]. The findings suggest that the APOE genotype should be taken into account for future research and treatment studies, in particular for EOAD.

Furthermore, our results showing decreased levels of Aβ42 in +LOAD compared with –EOAD are surprising. Although it has been well established that age comes with decreasing levels of Aβ42 [42], evidence showing Aβ42 differences between EOAD and LOAD has not been reported. It has been suggested that APOE4 could play an important role in Aβ42 accumulation in early stages of AD but not in more advanced stages of the disease [43, 44]. This also agrees with previous studies of our group, where no significant differences in Aβ42 levels were found on comparing EOAD and LOAD [45, 46]. However, previous studies have not classified AD participants with respect to AAO, and our previous studies did not classify patients according to their APOE genotype. Although the effect that APOE4 plays in Aβ42 as a function of age cannot be certain, it is an interesting point that has to be analyzed in future studies.

Despite these exciting findings, our study has several limitations. Stratifying patients by AAO and APOE status reduces group sizes for statistical contrasts. Most of our groups have reasonable sample sizes but clearly it would be important for future research with larger group sizes to confirm our results. Further, our ε4 sample included ε4/ε3 and ε4/ε2 with fewer participants presenting the ε4/ε4 genotype. This is of relevance as previous studies have suggested that carrying only one ε4 allele is not sufficient to induce macroscopic GM changes measured by VBM [31]. Also, the ε2 and ε3 alleles in heterozygous participants may have a differential effect in brain vulnerability compared to ε4 homozygous participants. Whilst the protective effect of ε2 for AD is overtaken by the ε4 allele in ε4/ε2, the relative risk for AD in ε4/ε4 is six times higher compared to heterozygous subjects [31]. This clearly suggests that our ε4 group could be in some way modulated by ε2 and ε3 alleles. Future studies should include homozygous e4 participants when comparing with non‐carriers to fully understand the structural and cognitive effects that APOE4 has in AD patients. The use of an uncorrected statistical threshold for our results can also be considered as a potential limitation. Given that our analysis was specific to the Papez circuit and considering our sample size, the probability of reporting false positives decreases substantially. Moreover, the same threshold has also been reported in previous VBM studies [47, 48]. Finally, on a cognitive level, not all patients completed all cognitive measures which might have influenced the cognitive–atrophy correlates. Nonetheless, there was greater cognitive–atrophy correlation in –LOAD which was the group with the lowest number of participants but overall better cognitive performance. Still, our study presents a first, preliminary evidence that the APOE genotype impacts different structural brain changes in EOAD and LOAD, which will inform future research and intervention studies of AD.

AUTHOR CONTRIBUTIONS

Gonzalo Forno contributed in conceptualization, formal analysis, methodology, validation, visualization, writing the original draft and writing—review and editing. Jose Contador, Agnès Pérez‐Millan, Nuria Guillen, Neus Falgàs, Jordi Sarto, Adrià Tort‐Merino, Magdalena Castellví, Beatriz Bosch, Guadalupe Fernández‐Villullas, Mircea Balasa, Anna Antonell and Roser Sala‐Llonch equally contributed in data curation, resources, validation, writing—review and editing. Raquel Sanchez‐Valle contributed in conceptualization, data curation, funding acquisition, investigation, resources, supervision, validation and writing—review and editing. Michael Hornberger contributed in conceptualization, investigation, methodology, supervision, validation, visualization, writing the original draft and writing—review and editing. Albert Llado contributed in conceptualization, data curation, funding acquisition, investigation, methodology, project administration, resources, supervision, validation, visualization, writing the original draft and writing—review and editing.

FUNDING INFORMATION

This work was supported by Instituto de Salud Carlos III (ISCIII) through the project PI19/00449 to Dr A Lladó and PI20/00448 to Dr R Sanchez‐Valle and co‐funded by the European Union, PERIS 2016‐2020 Departament de Salut de la Generalitat de Catalunya (SLT008 / 18/00061 to Dr A Lladó) and CERCA Programme/Generalitat de Catalunya. MH is funded by the Alzheimer’s Research UK, Wellcome Trust, Medical Research Council and BBSRC. NG received funding from a PFIS grant (FI20/00076). JS received funding from a PFIS grant (FI21/00015).

CONFLICT OF INTEREST

The authors declare that they have no conflict of interest.

Supporting information

Appendix S1:

Click here for additional data file. (2.6MB, docx)

ACKNOWLEDGEMENTS

The authors thank patients and their relatives for their participation in the research.

Forno G, Contador J, Pérez‐Millan A, et al. The APOE4 effect: structural brain differences in Alzheimer's disease according to the age at symptom onset. Eur J Neurol. 2023;30:597‐605. doi: 10.1111/ene.15657

Michael Hornberger and Albert Lladó share senior authorship.

DATA AVAILABILITY STATEMENT

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

REFERENCES

  • 1. Frisoni GB, Pievani M, Testa C, et al. The topography of grey matter involvement in early and late onset Alzheimer's disease. BrainBrain. 2007;130(Pt 3):720‐730. doi: 10.1093/brain/awl377 [DOI] [PubMed] [Google Scholar]
  • 2. Cho H, Jeon S, Kang SJ, et al. Longitudinal changes of cortical thickness in early‐ versus late‐onset Alzheimer's disease. Neurobiol AgingNeurobiol Aging. 2013;34(7):1921 e9‐1921 e15. doi: 10.1016/j.neurobiolaging.2013.01.004 [DOI] [PubMed] [Google Scholar]
  • 3. Cavedo E, Pievani M, Boccardi M, et al. Medial temporal atrophy in early and late‐onset Alzheimer's disease. Neurobiol AgingNeurobiol Aging. 2014;35(9):2004‐2012. doi: 10.1016/j.neurobiolaging.2014.03.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Kaiser NC, Melrose RJ, Liu C, et al. Neuropsychological and neuroimaging markers in early versus late‐onset Alzheimer's disease. Am J Alzheimers Dis Other DemenAm J Alzheimers Dis Other Demen. 2012;27(7):520‐529. doi: 10.1177/1533317512459798 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Falgas N, Allen IE, Spina S, et al. The severity of neuropsychiatric symptoms is higher in early‐onset than late‐onset Alzheimer's disease. Eur J NeurolEur J Neurol. 2022;29(4):957‐967. doi: 10.1111/ene.15203 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Koedam EL, Lauffer V, van der Vlies AE, van der Flier WM, Scheltens P, Pijnenburg YA. Early‐ versus late‐onset Alzheimer's disease: more than age alone. J Alzheimers DisJ Alzheimers Dis. 2010;19(4):1401‐1408. doi: 10.3233/JAD-2010-1337 [DOI] [PubMed] [Google Scholar]
  • 7. van der Flier WM, Pijnenburg YAL, Fox NC, Scheltens P. Early‐onset versus late‐onset Alzheimer's disease: the case of the missing APOE ε4 allele. Lancet NeurolLancet Neurol. 2011;10:280‐288. doi: 10.1016/S1474-4422(10)70306-9 [DOI] [PubMed] [Google Scholar]
  • 8. Strittmatter WJ, Roses AD. Apolipoprotein E and Alzheimer disease. Proc Natl Acad Sci U S AProc Natl Acad Sci U S A. 1995;92(11):4725‐4727. doi: 10.1073/pnas.92.11.4725 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Meyer MR, Tschanz JT, Norton MC, et al. APOE genotype predicts when—not whether—one is predisposed to develop Alzheimer disease. Nat GenetNat Genet. 1998;19(4):321‐322. doi: 10.1038/1206 [DOI] [PubMed] [Google Scholar]
  • 10. Pievani M, Galluzzi S, Thompson PM, Rasser PE, Bonetti M, Frisoni GB. APOE4 is associated with greater atrophy of the hippocampal formation in Alzheimer's disease. Neuroimage. 2011;55(3):909‐919. doi: 10.1016/j.neuroimage.2010.12.081 [DOI] [PubMed] [Google Scholar]
  • 11. Hirono N, Hashimoto M, Yasuda M, et al. The effect of APOE epsilon4 allele on cerebral glucose metabolism in AD is a function of age at onset. Neurology. 2002;58(5):743‐750. doi: 10.1212/wnl.58.5.743 [DOI] [PubMed] [Google Scholar]
  • 12. Sluimer JD, Vrenken H, Blankenstein MA, et al. Whole‐brain atrophy rate in Alzheimer disease: identifying fast progressors. Neurology. 2008;70(19 Pt 2):1836‐1841. doi: 10.1212/01.wnl.0000311446.61861.e3 [DOI] [PubMed] [Google Scholar]
  • 13. Kim J, Park S, Yoo H, et al. The impact of APOE varepsilon4 in Alzheimer's disease differs according to age. J Alzheimers DisJ Alzheimers Dis. 2018;61(4):1377‐1385. doi: 10.3233/JAD-170556 [DOI] [PubMed] [Google Scholar]
  • 14. Pievani M, Rasser PE, Galluzzi S, et al. Mapping the effect of APOE epsilon4 on gray matter loss in Alzheimer's disease in vivo. NeuroimageNeuroimage. 2009;45(4):1090‐1098. doi: 10.1016/j.neuroimage.2009.01.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Forno G, Llado A, Hornberger M. Going round in circles—the Papez circuit in Alzheimer's disease. Eur J NeurosciEur J Neurosci. 2021;54(10):7668‐7687. doi: 10.1111/ejn.15494 [DOI] [PubMed] [Google Scholar]
  • 16. Jack CR Jr, Bennett DA, Blennow K, et al. NIA‐AA research framework: toward a biological definition of Alzheimer's disease. Alzheimers DementAlzheimers Dement. 2018;14(4):535‐562. doi: 10.1016/j.jalz.2018.02.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. McKhann GM, Knopman DS, Chertkow H, et al. The diagnosis of dementia due to Alzheimer's disease: recommendations from the National Institute on Aging–Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers DementAlzheimers Dement. 2011;7(3):263‐269. doi: 10.1016/j.jalz.2011.03.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Antonell A, Tort‐Merino A, Rios J, et al. Synaptic, axonal damage and inflammatory cerebrospinal fluid biomarkers in neurodegenerative dementias. Alzheimers DementAlzheimers Dement. 2020;16(2):262‐272. doi: 10.1016/j.jalz.2019.09.001 [DOI] [PubMed] [Google Scholar]
  • 19. Ashburner J, Friston KJ. Voxel‐based morphometry—the methods. NeuroimageNeuroimage. 2000;11(6 Pt 1):805‐821. doi: 10.1006/nimg.2000.0582 [DOI] [PubMed] [Google Scholar]
  • 20. Rolls ET, Huang CC, Lin CP, Feng J, Joliot M. Automated anatomical labelling atlas 3. Neuroimage. 2020;206:116189. doi: 10.1016/j.neuroimage.2019.116189 [DOI] [PubMed] [Google Scholar]
  • 21. Hornberger M, Wong S, Tan R, et al. In vivo and post‐mortem memory circuit integrity in frontotemporal dementia and Alzheimer's disease. BrainBrain. 2012;135(Pt 10):3015‐3025. doi: 10.1093/brain/aws239 [DOI] [PubMed] [Google Scholar]
  • 22. Garcia‐Cordero I, Sedeno L, de la Fuente L, et al. Feeling, learning from and being aware of inner states: interoceptive dimensions in neurodegeneration and stroke. Philos Trans R Soc Lond B Biol SciPhilos Trans R Soc Lond B Biol Sci. 2016;371(1708):20160006. doi: 10.1098/rstb.2016.0006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Irish M, Piguet O, Hodges JR, Hornberger M. Common and unique gray matter correlates of episodic memory dysfunction in frontotemporal dementia and Alzheimer's disease. Hum Brain MappHum Brain Mapp. 2014;35(4):1422‐1435. doi: 10.1002/hbm.22263 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Grober E, Buschke H, Crystal H, Bang S, Dresner R. Screening for dementia by memory testing. NeurologyNeurology. 1988;38(6):900‐903. doi: 10.1212/wnl.38.6.900 [DOI] [PubMed] [Google Scholar]
  • 25. Pena‐Casanova J, Quintana‐Aparicio M, Quinones‐Ubeda S, et al. Spanish multicenter normative studies (NEURONORMA project): norms for the visual object and space perception battery‐abbreviated, and judgment of line orientation. Arch Clin Neuropsychol. 2009;24(4):355‐370. doi: 10.1093/arclin/acp040 [DOI] [PubMed] [Google Scholar]
  • 26. IBM Corp ; IBM SPSS statistics for windows, Version.25.0. 2017.
  • 27. Callen DJ, Black SE, Gao F, Caldwell CB, Szalai JP. Beyond the hippocampus: MRI volumetry confirms widespread limbic atrophy in AD. Neurology. 2001;57(9):1669‐1674. doi: 10.1212/wnl.57.9.1669 [DOI] [PubMed] [Google Scholar]
  • 28. Copenhaver BR, Rabin LA, Saykin AJ, et al. The fornix and mammillary bodies in older adults with Alzheimer's disease, mild cognitive impairment, and cognitive complaints: a volumetric MRI study. Psychiatry Res. 2006;147(2–3):93‐103. doi: 10.1016/j.pscychresns.2006.01.015 [DOI] [PubMed] [Google Scholar]
  • 29. Nestor PJ, Fryer TD, Hodges JR. Declarative memory impairments in Alzheimer's disease and semantic dementia. Neuroimage. 2006;30(3):1010‐1020. doi: 10.1016/j.neuroimage.2005.10.008 [DOI] [PubMed] [Google Scholar]
  • 30. Drzezga A, Riemenschneider M, Strassner B, et al. Cerebral glucose metabolism in patients with AD and different APOE genotypes. NeurologyNeurology. 2005;64(1):102‐107. doi: 10.1212/01.WNL.0000148478.39691.D3 [DOI] [PubMed] [Google Scholar]
  • 31. Crivello F, Lemaitre H, Dufouil C, et al. Effects of ApoE‐epsilon4 allele load and age on the rates of grey matter and hippocampal volumes loss in a longitudinal cohort of 1186 healthy elderly persons. NeuroimageNeuroimage. 2010;53(3):1064‐1069. doi: 10.1016/j.neuroimage.2009.12.116 [DOI] [PubMed] [Google Scholar]
  • 32. Coughlan G, Coutrot A, Khondoker M, Minihane AM, Spiers H, Hornberger M. Toward personalized cognitive diagnostics of at‐genetic‐risk Alzheimer's disease. Proc Natl Acad Sci U S A. 2019;116(19):9285‐9292. doi: 10.1073/pnas.1901600116 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Coughlan G, Zhukovsky P, Puthusseryppady V, et al. Functional connectivity between the entorhinal and posterior cingulate cortices underpins navigation discrepancies in at‐risk Alzheimer's disease. Neurobiol Aging. 2020;90:110‐118. doi: 10.1016/j.neurobiolaging.2020.02.007 [DOI] [PubMed] [Google Scholar]
  • 34. Gellersen HM, Coughlan G, Hornberger M, Simons JS. Memory precision of object‐location binding is unimpaired in APOE epsilon4‐carriers with spatial navigation deficits. Brain Commun. 2021;3(2):fcab087. doi: 10.1093/braincomms/fcab087 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Grober E, Veroff AE, Lipton RB. Temporal unfolding of declining episodic memory on the free and cued selective reminding test in the predementia phase of Alzheimer's disease: implications for clinical trials. Alzheimers Dement (Amst). 2018;10:161‐171. doi: 10.1016/j.dadm.2017.12.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Slachevsky A, Barraza P, Hornberger M, et al. Neuroanatomical comparison of the "word" and "picture" versions of the free and cued selective reminding test in Alzheimer's disease. J Alzheimers DisJ Alzheimers Dis. 2018;61(2):589‐600. doi: 10.3233/JAD-160973 [DOI] [PubMed] [Google Scholar]
  • 37. Li W, Antuono PG, Xie C, et al. Aberrant functional connectivity in Papez circuit correlates with memory performance in cognitively intact middle‐aged APOE4 carriers. CortexCortex. 2014;57:167‐176. doi: 10.1016/j.cortex.2014.04.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Novellino F, Lopez ME, Vaccaro MG, Miguel Y, Delgado ML, Maestu F. Association between hippocampus, thalamus, and caudate in mild cognitive impairment APOEepsilon4 carriers: a structural covariance MRI study. Front NeurolFront Neurol. 2019;10:1303. doi: 10.3389/fneur.2019.01303 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Machulda MM, Jones DT, Vemuri P, et al. Effect of APOE epsilon4 status on intrinsic network connectivity in cognitively normal elderly subjects. Arch NeurolArch Neurol. 2011;68(9):1131‐1136. doi: 10.1001/archneurol.2011.108 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Braak H, Braak E. Neuropathological staging of Alzheimer‐related changes. Acta Neuropathol. 1991;82(4):239‐259. doi: 10.1007/BF00308809 [DOI] [PubMed] [Google Scholar]
  • 41. Marra C, Bizzarro A, Daniele A, et al. Apolipoprotein E epsilon4 allele differently affects the patterns of neuropsychological presentation in early‐ and late‐onset Alzheimer's disease patients. Dement Geriatr Cogn Disord. 2004;18(2):125‐131. doi: 10.1159/000079191 [DOI] [PubMed] [Google Scholar]
  • 42. Morris JC, Roe CM, Xiong C, et al. APOE predicts amyloid‐beta but not tau Alzheimer pathology in cognitively normal aging. Ann Neurol. 2010;67(1):122‐131. doi: 10.1002/ana.21843 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Baek MS, Cho H, Lee HS, Lee JH, Ryu YH, Lyoo CH. Effect of APOE epsilon4 genotype on amyloid‐beta and tau accumulation in Alzheimer's disease. Alzheimers Res Ther. 2020;12(1):140. doi: 10.1186/s13195-020-00710-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44. Rowe CC, Ellis KA, Rimajova M, et al. Amyloid imaging results from the Australian imaging, biomarkers and lifestyle (AIBL) study of aging. Neurobiol AgingNeurobiol Aging. 2010;31(8):1275‐1283. doi: 10.1016/j.neurobiolaging.2010.04.007 [DOI] [PubMed] [Google Scholar]
  • 45. Contador J, Perez‐Millan A, Tort‐Merino A, et al. Longitudinal brain atrophy and CSF biomarkers in early‐onset Alzheimer's disease. Neuroimage Clin. 2021;32:102804. doi: 10.1016/j.nicl.2021.102804 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Falgas N, Sanchez‐Valle R, Bargallo N, et al. Hippocampal atrophy has limited usefulness as a diagnostic biomarker on the early onset Alzheimer's disease patients: a comparison between visual and quantitative assessment. Neuroimage Clin. 2019;23:101927. doi: 10.1016/j.nicl.2019.101927 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47. Baez S, Fittipaldi S, de la Fuente LA, et al. Empathy deficits and their behavioral, neuroanatomical, and functional connectivity correlates in smoked cocaine users. Prog Neuropsychopharmacol Biol Psychiatry. 2021;110:110328. doi: 10.1016/j.pnpbp.2021.110328 [DOI] [PubMed] [Google Scholar]
  • 48. Melloni M, Billeke P, Baez S, et al. Your perspective and my benefit: multiple lesion models of self–other integration strategies during social bargaining. Brain. 2016;139(11):3022‐3040. doi: 10.1093/brain/aww231 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix S1:

Click here for additional data file. (2.6MB, docx)

Data Availability Statement

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

Articles from European Journal of Neurology are provided here courtesy of Wiley

RESOURCES