Vitamin D Levels,APOE Allele,and MRI Volumetry Assessed by NeuroQuant in Norwegian Adults with Cognitive Symptoms

Abstract

Background:

Allele ɛ4 of the apolipoprotein (APOE ∈4) gene is the strongest known genetic risk factor for late-onset sporadic Alzheimer’s disease. A possible relationship between vitamin D and APOE is not yet clear.

Objective:

In this exploratory, cross-sectional study, we examined the association between serum levels of 25-hydroxyvitamin D [25(OH)D] and brain volumes and the associations of both serum levels of 25(OH)D and APOE polymorphism to brain volumes in 127 persons (mean age 66 years) with cognitive symptoms.

Methods:

All subjects were examined with fully automated software for MRI volumetry, NeuroQuant.

Results:

After adjustment for relevant covariates, higher serum 25(OH)D levels were associated with greater volumes of cortical gray matter on both left (p = 0.02) and right (p = 0.04) sides. When both 25(OH)D levels and APOE genotype were used as the main covariates, no significant associations were found between vitamin D level and brain volume in any of the 11 brain regions. In adjusted models, only homozygous but not heterozygous APOE ∈4 allele carriers had significantly larger inferior lateral ventricles (p = 0.003) and smaller hippocampal volume (p = 0.035) than those without ɛ4. Homozygous APOE ∈4 carriers also had significantly higher vitamin D levels (p = 0.009) compared to persons without the APOE ∈4 allele.

Conclusion:

Higher vitamin D levels might have a preserving effect on cortical grey matter volume.

Keywords

brain volumetry dementia executive function grey matter volume vitamin D

INTRODUCTION

As the life expectancy of the world population increases, dementia of all types is becoming one of the major and growing health challenges. The mech-anisms for the development of dementia are complex and not fully understood, involving among other factors, hypoxia, neuroinflammation, and oxi-dative stress. Therefore, a multidisciplinary approach is needed to improve the understanding of the development of dementia. Nutrition, including vitamin D, is an important area to thoroughly explore, especially as vitamin D deficiency is considered to be a global health problem, particularly among institutionalized elderly people [1]. There has been an increased inte-rest in the health effects of vitamin D, beyond the established knowledge of its beneficial effects on ost-eomalacia and rickets. In the last decades, it was discovered that nearly all cells in the human body contain receptors for vitamin D [2]. The two major forms of vitamin D are ergocalciferol, D2, and cholecalciferol, D3. The term vitamin D refers to both forms and the active metabolites calcidiol, 25-hydroxyvitamin D [25(OH)D], and calcitrol, 1,25-dihydroxyvitamin D [1,25(OH)2D], the latter being biologically active by binding to the vitamin D receptor (VDR). VDRs are found in tissues throughout the human body, in-cluding most regions of the brain, and are densely located in areas associated with memory and higher order cognition, such as the hippocampus, dentate and cingulate gyrus, and prefrontal cortex [3]. This discovery led to the hypothesis that vitamin D has neuroprotective effects [4, 5]. Cross-sectional and longitudinal studies indicate a positive relationship between vitamin D deficiency and cognitive dysfunction or deterioration and depressive symptoms [6 –9]. Executive functions seem to be more affected than memory [10 –13]. However, some studies failed to show a reliable relationship between vitamin D and cognitive decline [14 –16], questioning causality. Some intervention studies have shown better cognitive function in patients with AD receiving vitamin D [17] or in combination with memantine compared to memantine alone [18].

Morphological and volumetric brain changes in humans in relation to vitamin D levels have not been extensively studied. Previous neuroimaging studies found an association between vitamin D deficiency and higher grade of white matter lesions and de-creased grey matter volume [19 –22], but one prospective study could not find any association with the development of neurodegenerative imaging abnormalities [23]. A volumetric magnetic resonance im-aging (MRI) study of key brain regions, albeit with a small sample of 28 patients with memory complaints, found an association between higher vitamin D levels and greater brain volumes, including greater volumes of the amygdala, thalamus, and anterior cingulate gyrus [24].

Allele ɛ4 of the apolipoprotein (APOE) gene is the strongest known genetic risk factor for late-onset sporadic Alzheimer’s disease (AD) [25]. An individual can be a carrier of one or two APOE ∈4 alleles (hereinafter referred to as heterozygous or homozygous), or a non-carrier. It has been previously shown that carriers of at least one APOE ∈4 allele have a higher risk of atrophy of the hippocampus and amygdala compared to non-carriers [26, 27]. Two other major isoforms are APOE ∈2 and ∈3, the latter being the most common and considered to be neutral in disease development of AD. APOE ∈2 is relatively rare (estimated incidence 5%) and is considered to be a neuroprotective variant against AD [28]. The relationship between vitamin D and APOE is not yet clear. Intriguingly, some epidemiological data show that higher 25(OH)D concentrations might be particularly beneficial for memory function for those with two APOE ɛ4 alleles [29], whereas another study indicates that vitamin D deficiency presents a greater risk for APOE ∈4 non-carrier AD patients than for ∈4 carriers [30]. There are also data indicating that APOE ∈4 carriers have higher 25(OH)D concentrations, which is being interpreted as a potential evolutionary adaptation [31]. There is further more findings of the south-to-north gradient of the APOE ɛ4 frequencies [32], that may be explained by the possibility that APOE ∈4 carriers are less likely to develop vitamin D deficiency.

The development and use of MRI has enabled further mapping of the human brain and increased understanding of the course of neurodegenerative diseases. This modality provides great diagnostic detail and allows for volumetric measurement of key brain areas. The structural volumetric analyses can be performed manually, automatically, or as a combination of both. One of the more commonly used fully-automated software programs for MRI volumetry is NeuroQuant (NQ), which is FDA-approved. It has been found to be more sensitive for detecting brain atrophy than the standard manual volumetric measurement performed by radiologists [33].

As prior research is sparse regarding vitamin D and brain volumetry, we chose to perform an exploratory study, in which we investigated the associations between serum 25(OH)D, APOE polymorphisms, and brain volumetric measurements assessed by NQ in Norwegian adults presenting with cognitive symptoms.

MATERIALS AND METHODS

Subjects

The study was performed using data from the Norwegian registry of persons assessed for cognitive symptoms (NorCog). NorCog is a national quality and research registry that includes home-dwelling persons who are referred to a specialist for assessment of cognitive symptoms. Only patients with the ability to consent are included. The criteria for referral are self-reported cognitive symptoms or symptoms reported by a caregiver or medical staff. The NorCog registry includes a standard assessment package used by clinicians to document their clinical examination, cognitive testing, and interview with an informant (usually a caregiver or family member). Patients are referred for radiological imaging and laboratory analysis based on clinical indication. Diagnosis is determined after the evaluation is completed. Dementia diagnosis is based on ICD-10, whereas mild cognitive impairment (MCI) is based on the Winblad criteria [34]. Volumetric measurement using NQ is not a routine procedure in the standard assessment, but rather is part of a research project at Oslo University Hospital (OUH). The present study included persons registered in NorCog between 2013 and 2017 at OUH who had been assessed with NQ. Inclusion criteria were available NQ data and 25(OH)D serum level. In total, 127 persons were included.

Measurements and assessments

All 25(OH)D serum samples were analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS). Patients with 25(OH)D <50 nmol/l were considered to have vitamin D insufficiency and those with 25(OH)D <25 nmol/l were considered to have deficiency [35]. NQ measurements were perfo-rmed at OUH. T1-weighted 3D scanning was carried out on a 3 Tesla MRI scanner (General Electric Signa HDxt, Milwaukee, WI, USA) with the following parameters: repetition time (TR) = 7.8 ms; echo time (TE) = 2.90; field of view = 256×256×166; flip angle = 12°; slice thickness = 1.2 mm; voxel size =1.0×1.0×1.2 mm; and number of slices = 170.

NQ was updated in 2015 from version 1 to version 2. There are minor differences between the two versions that might be of significance [36], therefore the appropriate statistical analyses were performed to adjust for this. In the NQ versions used in this study, volumes for the following 11 brain regions were available: forebrain parenchyma, cortical gray matter, lateral ventricle, inferior lateral ventricle, hippocampus, amygdala, caudate nucleus, putamen, pallidum, thalamus, and cerebellum. Right and left side volumes were measured separately (in a total of 22 brain areas), and the volume of each region was divided by the total intracranial volume (% ICV) to correct for individual differences in head size.

Covariates

Age, sex, vascular risk factors (defined as one or more of the following: coronary or cerebrovascu-lar disease, hypertension or diabetes mellitus), smo-king habits (never smoker, former smoker or current smoker), and alcohol habits (lifetime abstainer/ infrequent drinker, light drinker or moderate/heavier drinker) were documented for all patients. It should be noted that the NorCog data do not differentiate between moderate and heavier alcohol consumption, as the highest level of alcohol consumption was 4–7 times per week. Data on education (in years) were available for 118 participants. Data on APOE polymorphism (individuals without APOE ∈4 allele, heterozygous individuals carrying one APOE ∈4 allele or homozygous for APOE ∈4 allele) were available for 115 participants.

Analysis

The sample was described with means and standard deviations (SD) or frequencies and percentages. As information about the APOE polymorphism was not available for all participants, the analyses were split into two parts. In Part I (N = 117), we estimated 22 models, one for each parameter for left and right side separately. Serum 25 (OH)D was the main covariate in this analysis. In Part II (N = 106), models were estimated with left and right side data merged together. Main covariates in Part II were serum 25 (OH)D and APOE polymorphism. First, unadjusted models were estimated. Next, all models were adjusted for pre-defined confounders (age, sex, vascular risk factors, smoking habits, alcohol habits, and education).

Additionally, all models were adjusted for NQ version (NQ1/NQ2) and for the period before and after May 2015. To correct for seasonal variations, serum 25(OH)D values were de-seasonalized in the following way. First, the seasonal function $S (t) = β_{1} + β \sin (\frac{2 π}{12} t) + β_{3} \cos (\frac{2 π}{12} t)$ (1)

where β₁, β₂, and β₃ are coefficients of the linear regression model and t defines the month of measurement (t = 1, . . . 12), was estimated. The resulting seasonal function values were then subtracted from the observed serum 25(OH)D values. All regression models were estimated on subsamples with no missing values on the covariates.

Independent-sample t-test was applied to assess differences in mean de-seasonalized vitamin D levels between persons without the APOE ∈4 allele and persons either heterozygous or homozygous for the APOE ∈4 allele (Analysis 1), and between persons without the APOE ∈4 allele and those who were homozygous APOE ∈4 carriers (Analysis 2).

Results with p-values < 0.05 were considered statistically significant. No adjustment for multiple tes-ting was performed as the study is exploratory. SPSS software (SPSS Inc., version 25.0) was used for data analysis.

Ethics

The study was approved by the Regional Committee for Medical and Health Research Ethics (reference 2016/888/REK Sør-Øst C). All included participants had given valid written consent for inclusion in the NorCog register.

RESULTS

A total of 127 persons were included. The dem-ographic characteristics of the study group are presented in Table 1. Their ages ranged from 38 to 86 years, with a mean of 66 years (SD 10). A slight maj-ority (53%) were men. Final diagnosis was subjective cognitive impairment (SCI) in 18%, MCI in 39%, and dementia in 39% of the participants. The remaining 4% received other diagnoses (e.g., Parkinson’s disease, multiple sclerosis). Mean serum 25(OH)D was 65 nmol/l (SD 21.5), ranging from 13 to 129 nmol/l. Median serum 25(OH)D was 64 nmol/l. The mean seasonal values of 25(OH)D are presented in Table 2. Serum 25(OH)D level indicated insufficiency (<50 nmol/l) in 19% of the participants, and only four (3%) also met the criterion for deficiency (<25 nmol/l). There was no significant difference in vitamin D levels between those with dementia and MCI (p = 0.86) or SCI (p = 0.71). For APOE genotype, 49 (38.6%) had the heterozygous APOE ∈4 allele polymorphism, 13 (10%) had the homozygous APOE ∈4 allele polymorphism, 53 (42%) had no APOE ∈4 allele, and 12 (9%) had no APOE genotype available.

Table 1

Descriptive characteristics of the study sample, N = 127

Characteristic	N (%)	Mean (SD)
Age, mean (SD)		66 (10)
Sex (women)	74 (47)
Education (y)¹		13 (4)
Smoking (never smoked)	66 (52)
Alcohol consumption (current/ former heavy drinker)²	20 (16)
Vascular risk factors	53 (41)
APOE∈4 polymorphism
Homozygous for APOE∈4 allele	13 (10)
Heterozygous for APOE∈4 allele	49 (39)
No APOE∈4 allele	53 (41)
No data available	12 (9)
Serum 25(OH)D in nmol/l		65 (21)
Sufficient (≥50 nmol/l)	103 (81)
Insufficiency (<50 nmol/l)	14 (19)
Deficient (<25 nmol/l)	4 (3)

¹N = 118. ²Consume alcohol 4 times a week or more, and includes former alcohol abuse.

Table 2

Mean seasonal values of 25(OH)D in nmol/l (study sample, N = 127, Age, mean 66 (SD 10))

Month	Mean value of 25(OH)D
January	61
February	60
March	60
April	62
May	64
June	67
July	69
August	70
September	70
October	69
November	66
December	64

Results of the linear mixed models are presented in Tables 3, 4. In Part I, serum 25(OH)D levels were not associated with brain volumes in unadjusted analyses. After adjustment for confounders, higher serum 25(OH)D levels were associated with greater volumes of cortical gray matter on both the left (p = 0.02) and the right (p = 0.04) sides.

Table 3

Part I: Associations between de-seasonalized serum 25(OH)D and brain volumetric (% of ICV) assessed through unadjusted and adjusted linear mixed models. N = 117 patients with no missing values on covariates.

N = 117	Unadjusted models		Adjusted models¹
	Regression coefficient	p	Regression coefficient	p
	(95% CI)		(95% CI)
Main covariate: de-seasonalized serum 25(OH)D
Forebrain parenchyma left	0.002 (–0.01; 0.02)	0.75	0.003 (–0.01; 0.02)	0.57
Forebrain parenchyma right	0.001 (–0.01; 0.02)	0.88	0.003 (–0.01; 0.02)	0.66
Cortical gray matter left	0.01 (–0.002; 0.02)	0.10	0.01 (0.002; 0.02)	0.02
Cortical gray matter right	0.01 (–0.003; 0.02)	0.14	0.01 (0.001; 0.02)	0.04
Lateral ventricle left	0.002 (–0.005; 0.01)	0.57	0.001 (–0.005; 0.01)	0.71
Lateral ventricle right	0.002 (–0.005; 0.01)	0.49	0.001 (–0.005; 0.01)	0.66
Inferior lateral ventricle left	0.0002 (–0.0003; 0.001)	0.43	0.00001 (–0.0003; 0.0004)	0.61
Inferior lateral ventricle right	0.0001 (–0.0003; 0.001)	0.61	0.0001 (–0.0003; 0.0005)	0.61
Hippocampus left	0.0001 (–0.0003; 0.0005)	0.57	0.0001 (–0.0002; 0.0004)	0.39
Hippocampus right	0.00002 (–0.0003; 0.0004)	0.91	0.00002 (–0.0003; 0.0003)	0.86
Amygdala left	–0.00003 (–0.0002; 0.0001)	0.76	–0.00003 (–0.0002; 0.0001)	0.66
Amygdala right	0.00003 (–0.0001; 0.0002)	0.69	0.00003 (–0.0001; 0.0001)	0.73
Caudate nucleus left	0.0003 (–0.0001; 0.001)	0.22	0.0002 (–0.0002; 0.001)	0.33
Caudate nucleus right	0.00001 (–0.0003; 0.0005)	0.88	–0.00004 (–0.0005; 0.0004)	0.85
Putamen left	–0.00003 (–0.0005; 0.005)	0.88	0.02 (–0.001; 0.04)	0.85
Putamen right	0.0002 (–0.0003; 0.001)	0.47	0.0001 (–0.0003; 0.001)	0.51
Pallidum left	0.00003 (–0.0001; 0.0001)	0.64	0.00005 (–0.0001; 0.0002)	0.41
Pallidum right	0.00003 (–0.0001; 0.0001)	0.55	0.00003 (–0.0001; 0.0001)	0.61
Thalamus left	0.0004 (–0.0001; 0.001)	0.13	0.0005 (–0.0001; 0.001)	0.08
Thalamus right	0.0003 (–0.0001; 0.001)	0.16	0.0004 (–0.0001; 0.001)	0.13
Cerebellum left	0.001 (–0.002; 0.004)	0.69	0.001 (–0.002; 0.003)	0.73
Cerebellum right	0.0001 (–0.003; 0.003)	0.93	–0.00002 (–0.003; 0.003)	0.99

¹All numbers are adjusted for NQ-version and time before/after May 2015. Adjusted models include the confounders of sex, age, education, smoking habits, alcohol consumption, and vascular risk factors.

Table 4

Part II: Associations of both de-seasonalized serum 25(OH)D and APOE ∈4 allele polymorphism with brain volumetric (% of ICV), assessed through unadjusted and adjusted linear mixed models. N = 106 patients with no missing values on covariates.

N = 106	Unadjusted model		Adjusted model¹
	Regression coefficient	p	Regression coefficient	p
	(95% CI)		(95% CI)
Forebrain parenchyma²
25 (OH)D	0.001 (–0.02; 0.02)	0.88	0.004 (–0.01; 0.02)	0.54
No APOE∈4	0		0
Heterozygous APOEɛ4	–0.40 (–1.08; 0.29)	0.25	–0.04 (–0.63; 0.56)	0.91
Homozygous APOEɛ4	–0.83 (–1.87; 0.22)	0.12	0.09 (–0.83; 1.01)	0.85
Cortical gray matter²
25 (OH)D	0.01 (–0.01; 0.02)	0.25	0.01 (–0.002; 0.02)	0.10
No APOE∈4	0		0
Heterozygous APOEɛ4	–0.19 (–0.78; 0.40)	0.52	0.02 (–0.53; 0.57)	0.93
Homozygous APOEɛ4	–0.34 (–1.24; 0.56)	0.46	0.15 (–0.70; 1.01)	0.73
Lateral ventricle²
25 (OH)D	0.004 (–0.004; 0.01)	0.31	0.001 (–0.006; 0.01)	0.69
No APOE∈4	0		0
Heterozygous APOEɛ4	0.15 (–0.17; 0.48)	0.36	–0.002 (–0.32; 0.31)	0.99
Homozygous APOEɛ4	0.76 (0.26; 1.25)	0.003	0.44 (–0.06; 0.93)	0.08
Inferior lateral ventricle²
25 (OH)D	0.0002 (–0.0002; 0.001)	0.30	0.0004 (–0.0003; 0.0004)	0.82
No APOE∈4	0		0
Heterozygous APOEɛ4	0.01 (–0.004; 0.03)	0.13	0.01 (–0.01; 0.02)	0.37
Homozygous APOEɛ4	0.06 (0.03; 0.09)	<0.001	0.04 (0.01; 0.06)	0.003
Hippocampus²
25 (OH)D	–0.00001 (–0.0004; 0.0003)	0.72	–0.00001 (–0.0002; 0.0003)	0.67
No APOE∈4	0		0
Heterozygous APOEɛ4	–0.02 (–0.03; 0.0003)	0.06	–0.01 (–0.02; 0.001)	0.08
Homozygous APOEɛ4	–0.04 (–0.06; –0.01)	0.003	–0.02 (–0.04; –0.002)	0.035
Amygdala²
25 (OH)D	–0.00001 (–0.0002; 0.0001)	0.80	–0.00001 (–0.0001; 0.0001)	0.86
No APOE∈4	0		0
Heterozygous APOEɛ4	–0.005 (–0.01; 0.002)	0.14	–0.003 (–0.01; 0.003)	0.41
Homozygous APOEɛ4	–0.02 (–0.02; –0.005)	0.003	–0.01 (–0.02; 0.001)	0.07
Caudate nucleus²
25 (OH)D	0.0002 (–0.0002; 0.001)	0.39	0.00004 (–0.0004; 0.0004)	0.84
No APOE∈4	0		0
Heterozygous APOEɛ4	0.01 (–0.004; 0.03)	0.13	0.01 (–0.01; 0.02)	0.53
Homozygous APOEɛ4	0.03 (0.003; 0.06)	0.03	0.02 (–0.01; 0.04)	0.22
Putamen²
25 (OH)D	–0.00002 (–0.0005; 0.0004)	0.94	–0.000002 (–0.0004; 0.0004)	0.99
No APOE∈4	0		0
Heterozygous APOEɛ4	0.01 (–0.01; 0.03)	0.26	0.01 (–0.01; 0.03)	0.37
Homozygous APOEɛ4	–0.003 (–0.03; 0.03)	0.86	0.004 (–0.03; 0.03)	0.77
Pallidum²
25 (OH)D	0.00001 (–0.0001; 0.0001)	0.86	0.00004 (–0.0001; 0.0001)	0.51
No APOE∈4	0		0
Heterozygous APOEɛ4	–0.004 (–0.01; 0.001)	0.11	–0.003 (–0.01; 0.002)	0.28
Homozygous APOEɛ4	–0.01 (–0.01; 0.001)	0.11	–0.003 (–0.01; 0.005)	0.44
Thalamus²
25 (OH)D	0.0004 (–0.0001; 0.001)	0.11	0.0004 (–0.0001; 0.001)	0.10
No APOE∈4	0		0
Heterozygous APOEɛ4	0.01 (–0.01; 0.03)	0.42	0.01 (–0.01; 0.03)	0.41
Homozygous APOEɛ4	0.02 (–0.01; 0.06)	0.18	0.03 (–0.002; 0.07)	0.07
Cerebellum²
25 (OH)D	0.00001 (–0.003; 0.003)	0.10	–0.00004 (–0.003; 0.003)	0.98
No APOE∈4	0		0
Heterozygous APOEɛ4	–0.01 (–0.15; 0.13)	0.94	0.05 (–0.09; 0.19)	0.51
Homozygous APOEɛ4	–0.07 (–0.28; 0.14)	0.52	–0.005 (–0.23; 0.22)	0.96

¹All numbers are adjusted for NQ version and time before/after May 2015. Adjusted models include the confounders of sex, age, education, smoking habits, alcohol consumption, and vascular risk factors. ²Average of left and right side.

In Part II, there were no significant associations between vitamin D level and brain volume in any of the 11 brain regions, in either unadjusted or adj-usted analyses. In the unadjusted models, persons with the homozygous APOE ∈4 allele polymorphism had significantly larger lateral ventricles (p = 0.003), inferior lateral ventricles (p < 0.001), and caudate nucleus volumes (p = 0.03), and significantly smaller hippocampus (p = 0.003) and amygdala volumes (p =0.003) than those without ɛ4. In the adjusted model, homozygous APOE ∈4 allele carriers had significantly larger inferior lateral ventricles (p = 0.003) and smaller hippocampal volume (p = 0.035) than those without ɛ4. We did not find any significant differences in the brain volumes of persons with the heterozygous APOE ∈4 allele polymorphism compared to those without the ɛ4 allele.

Table 5 shows the associations between mean de-seasonalized vitamin D levels and APOE ∈4 allele carrier status. There were no significant differences between persons without the APOE ∈4 allele and persons with either the heterozygous or homozygous APOE ∈4 allele polymorphism in mean vitamin D values (Analysis 1, p = 0.10). Homozygous APOE ∈4 carriers had significantly higher vitamin D levels (p = 0.009) compared to persons without the APOE ∈4 allele (Analysis 2).

Table 5

Differences in de-seasonalized serum 25(OH)D (nmol/l) values by APOE ∈4 polymorphism

Analysis 1		Heterozygous or homozygous APOE∈4	p
De-seasonalized serum	–3 (22.5)¹	3 (19.5)	0.10
Vitamin D, mean (SD)
Analysis 2		Homozygous APOE∈4
De-seasonalized serum	–3 (22.5)¹	9 (12)	0.009
Vitamin D, mean (SD)

¹No APOE ∈4 allele.

DISCUSSION

In this exploratory, cross-sectional study, we examined the associations between serum levels of 25(OH)D and brain volumes in a cohort of persons with cognitive symptoms, the majority diagnosed with dementia or MCI. Vitamin D insufficiency (de-fined as 25(OH)D <50 nmol/l) was present in 18.9% of the cohort, 7% had vitamin D levels <30 nmol/l and 3% <25 nmol/l. This is lower than the prevalence found in previous Norwegian population studies [37]. The reason for this decrease might be awareness around the importance of vitamin D in the general population in recent years and increased use of vitamin D supplements. Comparably in the Framingham Heart Study, only 8% of participant had vitamin D levels <25 nmol/l [13].

Our results show an association between higher serum levels of vitamin D and increased volumes of both left and right cortical gray matter. We did not find an association between vitamin D levels and the volume of other brain areas. Previous studies in this area are sparse. Berg and colleagues found that in patients with psychosis, vitamin D levels were positively associated with grey matter volume, but not with the volumes of other brain areas [21], a finding similar to our results, but in a different patient population. In contrast, our results did not confirm previous findings of vitamin D depletion being associated with lower brain volume, specifically larger lateral ventricles [38]. A recent study by Croll et al. found that vitamin D levels <30 nmol/l were associated with smaller brain tissue volume, but the association was not found for levels between 30–50 nmol/l (defined as insufficiency) [39]. Other studies found reduced hippocampal volume in vitamin D deficient persons, also here defined as persons with vitamin D levels < 25 or 30 nmol/l [13, 40]. These findings are not confirmed in our study. A possible explanation for this is a small number of persons with vitamin D insufficiency and deficiency in our cohort. We could not reproduce the findings of Hooshmand and colleagues, who found an association between increased 25(OH)D and larger total brain volumes, and amygdala, thalamus, and anterior cingulate gyrus volumes [24].

Vitamin D levels >30 nmol/l might be of critical importance for brain health. Our results might indicate that even higher levels are necessary for optimal function of cortical grey matter.

Cortical grey matter volume has been shown to be associated with executive functions [41], a set of cognitive functions essential for adaptation to new stimuli, and responsible for cognitive tasks, such as working memory, information updating, divided att-ention, and inhibition. Our earlier research showed an association between vitamin D levels and executive functions in persons with cognitive symptoms [10], which has also been found in other studies [11]. Since executive functions and attention are closely related to the attention brain network, which is localized only in the cortex [42], our results might indicate positive correlations between vitamin D concentration, cortical volume, and attention.

Several possible biological mechanisms might explain the effect of vitamin D on the central nervous system and cortical grey matter. Activated vitamin D is a ligand of VDRs which are heavily distributed in the superficial/associative layer of the cortex, am-ong other brain regions [3]. Studies have found and increased risk of AD in persons with certain VDR polymorphisms [43]. Vitamin D deficiency is further suspected to change brain microenvironment and induce amyloid aggregation [44] and vitamin D is shown to have a neuroprotective and anti-inflam-matory action on brain cells [45, 46].

We found no associations between vitamin D and any brain volumes when APOE polymorphism was used as a second main covariate (Part II of the analysis), although there was a trend toward larger grey matter volume in persons with higher vitamin D levels. Data on APOE polymorphism and education were not available for all persons in the cohort, thus reducing the sample size. One possible explanation for the lack of significant association in Part II might be that the sample size was too small.

We found significantly higher vitamin D levels in the group with the homozygous APOE ∈4 polymorphism compared to the group with no APOE ∈4 allele, supporting the hypothesis that the APOE ɛ4 allele protects against vitamin D deficiency by increasing intestinal absorption and renal re-uptake of vitamin D [47]. This might be an evolutionary phenomenon that could explain the survival of APOE ɛ4 allele and the higher prevalence of the APOE ɛ4 allele in high-latitude cold environments with less sun exposure than in low-latitude hot environments [48].

There is still an ongoing debate about the optimal serum levels of vitamin D. As levels of 25(OH)D <25 nmol/l are known to cause rickets and osteomalacia [49], this level is thus used in several guidelines as the level that constitutes a deficiency. In 2003, the World Health Organization (WHO) defined vitamin D insufficiency as 25(OH)D <50 nmol/l, based on the levels needed to prevent osteoporosis and muscle weakness [50]. The National Academy of Medicine (formerly the Institute of Medicine, IOM) defined in 2011 the levels of 25(OH)D 30–50 nmol/l as the normal range for adequate exposure to vitamin D to maintain bone health [51]. Optimal vitamin D levels for extra-skeletal functions, including brain health, are under still under discussion. The ‘Endocrine Society Clinical Practice Guideline’ from 2011 defined the vitamin D deficiency as levels of 25(OH)D <50 nmol/l and insufficiency as 25(OH)D <75 nmol [52]. The last years there has been an intensive scientific debate on the differences in the recommendations regarding vitamin D requirements from the IOM report and the Endocrine Society Clinical Practice guideline [53] which also affect the evaluation of how large part of the world population is vitamin D deficient. There are still ongoing randomized controlled trials examining the evidence for non-skeletal health effects of higher circulating 25(OH)D (in ex-cess of 50 nmol/L for bone health) that might change the nutritional recommendations in the future. The recent large randomized controlled trial, VITAL, did not find lower incidence of invasive cancer or cardiovascular events than placebo in persons that used supplementation with vitamin D [54].

The current Norwegian recommendation consider 25(OH)D <50 nmol/l as vitamin D insufficiency, in lack of evidence of health benefits of higher serum levels and out of concern of adverse health effects in overdoses of vitamin D [35]. It is due to note that vitamin D toxicity (clinically characterized by hypercalciuria and hypercalcemia) is extremely rare in literature and it is estimated that 25(OH)D concentrations must exceed 375 nmol/l, often in combination with other factors, such as calcium intake, for risk of developing hypercalcemia and clinical toxicity [55].

Several factors might influence the serum levels of 25(OH)D, as the sun exposure, season, diet [56], age, body mass index [57], and skin pigmentation. The Norwegian national council for nutrition recommends daily intake of 10μg vitamin D for young and adults and daily intake of 20μg vitamin D for persons over 75 years of age, on the assumption of minimal sunlight exposure. For healthy people staying outdoors, vitamin D supplements are not recommended in the summer [58]. We do not have the accurate data on vitamin D supplements and sun exposure in our study group, but the general intake of fish and od-liver oil in Norway is high [59]. All the participants were Caucasian. Vitamin D insufficiency and deficiency were not highly prevalent in our study cohort (19% had 25(OH)D <50 nmol/l and only 3% <25 nmol/l). It is therefore possible that the observed trends would be stronger in a population with higher prevalence of vitamin D insufficiency and/or deficiency. On the other hand, it should be noted that the observed association between higher vitamin D levels and larger cortical grey matter volume was present even in a population that mostly (81%) had sufficient vitamin D levels. This might mean that the optimal serum levels of 25(OH)D for brain health should be higher than current recommendations, but the positive findings need to be confirmed in further studies with sufficient power.

Allele ɛ4 of the APOE gene is the strongest known genetic risk factor for late-onset sporadic Alzheimer’s disease [25] and has been shown to have an effect on hippocampal volume asymmetry [60]. In our sample, homozygous APOE ɛ4 carriers had significantly larger inferior lateral ventricles and lower hippocampal volumes as compared to those with no APOE ɛ4 allele, but there was no statistically significant association with other brain volumes. No sig-nificant differences in brain volumes were found between heterozygous APOE ɛ4 carriers and those with no APOE ɛ4 allele. Several previous studies that found the association between hippocampal atrophy and APOE ɛ4 carriers did not distinguish between homozygous and heterozygous carriers of the polymorphism [27 , 62]. We also confirmed earlier findings of higher vitamin D levels in homozygous APOE ɛ4 carriers compared to APOE ɛ4 non-carriers, suggesting an evolutionary adaptation mechanism, but the association was not found when both heterozygous and homozygous APOE ɛ4 carriers were combined into one group for analysis. Our results indicate that there might be an important clinical distinction between carriers of one and two APOE ɛ4 alleles that should be considered in future studies [63]. It is also important to note that our study did not exclude carriers of the APOE ɛ4/ɛ2 variant, which could be considered in further studies, as this variant might have a neutral effect compared to other heterozygous APOE ɛ4 carriers.

Study strengths

The strength of this study is the large number of participants with NQ data and available information on APOE polymorphisms, comorbidity, education, smoking and alcohol use, allowing the models to be adjusted for known cofounders that might influence both brain volumes and vitamin D levels.

Study limitations

In the regression models for Part II of the analysis, complete covariate data were available for only 83% of the total study sample (compared to 92% in Part I), thereby reducing the number of participants in the analysis and possibly introducing bias.

Vitamin D was measured at the time of the initial medical examination. Patients were then referred for NQ examination, which was completed an average of 1–3 months after the medical examination. In a few cases, the time between the measurement of vitamin D and NQ exam was longer than 6 months. Thus, vitamin D and NQ were not taken at the same time. Variation in vitamin D levels, not accounted for by season, could have occurred, potentially weakening observed associations. It is also worth noting that the data did not differentiate between moderate and heavy use of alcohol, which may limit the usefulness of this measure.

This being an exploratory cross-sectional study, it is not possible to draw any conclusion about causality, as both lower vitamin D levels and smaller volumes of grey matter could be markers of poorer health. Even though the potentially confounding variables identified in previous studies were corrected for, residual confounding cannot be ruled out.

Conclusion

The present results indicate that vitamin D levels might have an influence on cortical grey matter volume and thus explain its effect on executive functions. This study supports previous findings of significantly decreased hippocampal volumes and better vitamin D status in homozygous APOE ɛ4 carriers, but not in heterozygous APOE ɛ4 carriers. Sufficient power and distinction between types of APOE ɛ4 carriers are recommended in future studies.

Footnotes

ACKNOWLEDGMENTS

We want to acknowledge the Academic Board of the Norwegian register of persons assessed for cognitive symptoms (NorCog) for providing access to the patient data described in this study.

Authors’ disclosures available online ().

References

Hilger

, Friedel

, Herr

, Rausch

, Roos

, Wahl

, Pierroz

, Weber

, Hoffmann

(2014) A systematic review of vitamin D status in populations worldwide. Br J Nutr 111, 23–45.

Stumpf

, Sar

, Reid

, Tanaka

, DeLuca

(1979) Target cells for 1,25-dihydroxyvitamin D3 in intestinal tract, stomach, kidney, skin, pituitary, and parathyroid. Science 206, 1188–1190.

Eyles

, Smith

, Kinobe

, Hewison

, McGrath

(2005) Distribution of the vitamin D receptor and 1 alpha-hydroxylase in human brain. J Chem Neuroanat 29, 21–30.

Siegel

, Malkowitz

, Moskovits

, Christakos

(1984) Administration of 1,25-dihydroxyvitamin D3 results in the elevation of hippocampal seizure threshold levels in rats. Brain Res 298, 125–129.

Stumpf

, O’Brien

(1987) 1,25 (OH)2 vitamin D3 sites of action in the brain. An autoradiographic study. Histochemistry 87, 393–406.

Annweiler

, Montero-Odasso

, Hachinski

, Seshadri

, Bartha

, Beauchet

(2013) Vitamin D concentration and lateral cerebral ventricle volume in older adults. Mol Nutr Food Res 57, 267–276.

Anglin

, Samaan

, Walter

, McDonald

(2013) Vitamin D deficiency and depression in adults: Systematic review and meta-analysis. Br J Psychiatry 202, 100–107.

Littlejohns

, Henley

, Lang

, Annweiler

, Beauchet

, Chaves

, Fried

, Kestenbaum

, Kuller

, Langa

, Lopez

, Kos

, Soni

, Llewellyn

(2014) Vitamin D and the risk of dementia and Alzheimer disease. Neurology 83, 920–928.

Sommer

, Griebler

, Kien

, Auer

, Klerings

, Hammer

, Holzer

, Gartlehner

(2017) Vitamin D deficiency as a risk factor for dementia: A systematic review and meta-analysis. BMC Geriatr 17, 16.

10.

Zugic Soares

, Pettersen

, Saltyte Benth

, Knapskog

, Selbaek

, Bogdanovic

(2019) Higher vitamin D levels are associated with better attentional functions: Data from the NorCog Register. J Nutr Health Aging 23, 725–731.

11.

Annweiler

, Montero-Odasso

, Llewellyn

, Richard-Devantoy

, Duque

, Beauchet

(2013) Meta-analysis of memory and executive dysfunctions in relation to vitamin D. J Alzheimers Dis 37, 147–171.

12.

Jorde

, Mathiesen

, Rogne

, Wilsgaard

, Kjaergaard

, Grimnes

, Schirmer

(2015) Vitamin D and cognitive function: The Tromso Study. J Neurol Sci 355, 155–161.

13.

Karakis

, Pase

, Beiser

, Booth

, Jacques

, Rogers

, DeCarli

, Vasan

, Wang

, Himali

, Annweiler

, Seshadri

(2016) Association of serum vitamin D with the risk of incident dementia and subclinical indices of brain aging: The Framingham Heart Study. J Alzheimers Dis 51, 451–461.

14.

Granic

, Hill

, Kirkwood

, Davies

, Collerton

, Martin-Ruiz

, von Zglinicki

, Saxby

, Wesnes

, Collerton

, Mathers

, Jagger

(2015) Serum 25-hydroxyvitamin D and cognitive decline in the very old: The Newcastle 85+Study. Eur J Neurol 22, 106–107, e106-107.

15.

Balion

, Griffith

, Strifler

, Henderson

, Patterson

, Heckman

, Llewellyn

, Raina

(2012) Vitamin D, cognition, and dementia: A systematic review and meta-analysis. Neurology 79, 1397–1405.

16.

Etgen

, Sander

, Bickel

, Sander

, Forstl

(2012) Vitamin D deficiency, cognitive impairment and dementia: A systematic review and meta-analysis. Dement Geriatr Cogn Disord 33, 297–305.

17.

Gangwar

, Rawat

, Tiwari

, Narayan

, Tiwari

(2015) Role of vitamin-D in the prevention and treatment of Alzheimer’s disease. Indian J Physiol Pharmacol 59, 94–99.

18.

Annweiler

, Herrmann

, Fantino

, Brugg

, Beauchet

(2012) Effectiveness of the combination of memantine plus vitamin D on cognition in patients with Alzheimer disease: A pre-post pilot study. Cogn Behav Neurol 25, 121–127.

19.

Annweiler

, Annweiler

, Bartha

, Herrmann

, Camicioli

, Beauchet

(2014) Vitamin D and white matter abnormalities in older adults: A cross-sectional neuroimaging study. Eur J Neurol 21, 1436–e1495.

20.

Annweiler

, Bartha

, Goncalves

, Karras

, Millet

, Feron

, Beauchet

(2015) Vitamin D-related changes in intracranial volume in older adults: A quantitative neuroimaging study. Maturitas 80, 312–317.

21.

Berg

, Jorgensen

, Nerhus

, Athanasiu

, Popejoy

, Bettella

, Norbom

LCB

, Gurholt

, Dahl

, Andreassen

, Djurovic

, Agartz

, Melle

(2018) Vitamin D levels, brain volume, and genetic architecture in patients with psychosis. PLoS One 13, e0200250.

22.

Brouwer-Brolsma

, van der Zwaluw

, van Wijngaarden

, Dhonukshe-Rutten

, in ‘t Veld

, Feskens

, Smeets

, Kessels

, van de Rest

, de Groot

(2015) Higher serum 25-hydroxyvitamin D and lower plasma glucose are associated with larger gray matter volume but not with white matter or total brain volume in dutch community-dwelling older adults. J Nutr 145, 1817–1823.

23.

Littlejohns

, Kos

, Henley

, Lang

, Annweiler

, Beauchet

, Chaves

, Kestenbaum

, Kuller

, Langa

, Lopez

, Llewellyn

(2016) Vitamin D and risk of neuroimaging abnormalities. PLoS One 11, e0154896.

24.

Hooshmand

, Lokk

, Solomon

, Mangialasche

, Miralbell

, Spulber

, Annerbo

, Andreasen

, Winblad

, Cedazo-Minguez

, Wahlund

, Kivipelto

(2014) Vitamin D in relation to cognitive impairment, cerebrospinal fluid biomarkers, and brain volumes. J Gerontol A Biol Sci Med Sci 69, 1132–1138.

25.

Samieri

, Proust-Lima

, M

, Okereke

, Amariglio

, Sperling

, Rentz

, Grodstein

(2014) Subjective cognitive concerns, episodic memory, and the APOE epsilon4 allele. Alzheimers Dement 10, 752–759 e751.

26.

Honea

, Vidoni

, Harsha

, Burns

(2009) Impact of APOE on the healthy aging brain: A voxel-based MRI and DTI study. J Alzheimers Dis 18, 553–564.

27.

O’Dwyer

, Lamberton

, Matura

, Tanner

, Scheibe

, Miller

, Rujescu

, Prvulovic

, Hampel

(2012) Reduced hippocampal volume in healthy young ApoE4 carriers: An MRI study. PLoS One 7, e48895.

28.

, Zhao

(2016) ApoE2 and Alzheimer’s disease: Time to take a closer look. Neural Regen Res 11, 412–413.

29.

Maddock

, Cavadino

, Power

, Hypponen

(2015) 25-hydroxyvitamin D, APOE varepsilon4 genotype and cognitive function: Findings from the 1958 British birth cohort. Eur J Clin Nutr 69, 505–508.

30.

Dursun

, Alaylioglu

, Bilgic

, Hanagasi

, Lohmann

, Atasoy

, Candas

, Araz

, Onal

, Gurvit

, Yilmazer

, Gezen-Ak

(2016) Vitamin D deficiency might pose a greater risk for ApoEvarepsilon4 non-carrier Alzheimer’s disease patients. Neurol Sci 37, 1633–1643.

31.

Egert

, Rimbach

, Huebbe

(2012) ApoE genotype: From geographic distribution to function and responsiveness to dietary factors. Proc Nutr Soc 71, 410–424.

32.

, Qin

, Jing

, Lu

, Hu

, Du

(2011) Does the geographical gradient of ApoE4 allele exist in China? A systemic comparison among multiple Chinese populations. Mol Biol Rep 38, 489–494.

33.

Ross

, Ochs

, Seabaugh

, Shrader

, Alzheimer’s Disease Neuroimaging Initiative (2013) Man versus machine: Comparison of radiologists’ interpretations and NeuroQuant(R) volumetric analyses of brain MRIs in patients with traumatic brain injury. J Neuropsychiatry Clin Neurosci 25, 32–39.

34.

Winblad

, Palmer

, Kivipelto

, Jelic

, Fratiglioni

, Wahlund

, Nordberg

, Backman

, Albert

, Almkvist

, Arai

, Basun

, Blennow

, de Leon

, DeCarli

, Erkinjuntti

, Giacobini

, Graff

, Hardy

, Jack

, Jorm

, Ritchie

, van Duijn

, Visser

, Petersen

(2004) Mild cognitive impairment–beyond controversies, towards a consensus: Report of the International Working Group on Mild Cognitive Impairment. J Intern Med 256, 240–246.

35.

Holvik

, Meyer

, Madar

, Brustad

(2019) High-dosage vitamin D supplements are unnecessary. Tidsskr Nor Laegeforen 139.

36.

Stelmokas

, Yassay

, Giordani

, Dodge

, Dinov

, Bhaumik

, Sathian

, Hampstead

(2017) Translational MRI volumetry with NeuroQuant: Effects of version and normative data on relationships with memory performance in healthy older adults and patients with mild cognitive impairment. J Alzheimers Dis 60, 1499–1510.

37.

Larose

, Chen

, Camargo

Jr. , Langhammer

, Romundstad

, Mai

(2014) Factors associated with vitamin D deficiency in a Norwegian population: The HUNT Study. J Epidemiol Community Health 68, 165–170.

38.

Annweiler

, Annweiler

, Montero-Odasso

, Bartha

, Beauchet

(2014) Vitamin D and brain volumetric changes: Systematic review and meta-analysis. Maturitas 78, 30–39.

39.

Croll

, Boelens

, Vernooij

, van de Rest

, Zillikens

, Ikram

, Voortman

(2020) Associations of vitamin D deficiency with MRI markers of brain health in a community sample. Clin Nutr, doi: 10.1016/j.clnu.2020.04.027.

40.

Al-Amin

, Bradford

, Sullivan

RKP

, Kurniawan

, Moon

, Han

, Zalesky

, Burne

THJ

(2019) Vitamin D deficiency is associated with reduced hippocampal volume and disrupted structural connectivity in patients with mild cognitive impairment. Hum Brain Mapp 40, 394–406.

41.

Manard

, Bahri

, Salmon

, Collette

(2016) Relationship between grey matter integrity and executive abilities in aging. Brain Res 1642, 562–580.

42.

Vossel

, Geng

, Fink

(2014) Dorsal and ventral attention systems: Distinct neural circuits but collaborative roles. Neuroscientist 20, 150–159.

43.

Gezen-Ak

, Dursun

, Bilgic

, Hanagasi

, Ertan

, Gurvit

, Emre

, Eker

, Ulutin

, Uysal

, Yilmazer

(2012) Vitamin D receptor gene haplotype is associated with late-onset Alzheimer’s disease. Tohoku J Exp Med 228, 189–196.

44.

Dursun

, Gezen-Ak

(2019) Vitamin D basis of Alzheimer’s disease: From genetics to biomarkers. Hormones (Athens) 18, 7–15.

45.

Brewer

, Thibault

, Chen

, Langub

, Landfield

, Porter

(2001) Vitamin D hormone confers neuroprotection in parallel with downregulation of L-type calcium channel expression in hippocampal neurons. J Neurosci 21, 98–108.

46.

Moore

, Piazza

, McCartney

, Lynch

(2005) Evidence that vitamin D3 reverses age-related inflammatory changes in the rat hippocampus. Biochem Soc Trans 33, 573–577.

47.

Huebbe

, Nebel

, Siegert

, Moehring

, Boesch-Saa-datmandi

, Most

, Pallauf

, Egert

, Muller

, Schreiber

, Nothlings

, Rimbach

(2011) APOE epsilon4 is associated with higher vitamin D levels in targeted replacement mice and humans. FASEB J 25, 3262–3270.

48.

Eisenberg

, Kuzawa

, Hayes

(2010) Worldwide allele frequencies of the human apolipoprotein E gene: Climate, local adaptations, and evolutionary history. Am J Phys Anthropol 143, 100–111.

49.

Holick

(2007) Vitamin D deficiency. N Engl J Med 357, 266–281.

50.

WHO Scientific Group on the Prevention and Management of Osteoporosis (2000: Geneva, Switzerland) (2003) Prevention and management of osteoporosis: Report of a WHO scientific group. World Health Organization. https://apps.who.int/iris/handle/10665/42841.

51.

Atkinson

(2011) The new dietary reference intakes from the Institute of Medicine for calcium and vitamin D. Perspect Infirm 8, 5.

52.

Holick

, Binkley

, Bischoff-Ferrari

, Gordon

, Hanley

, Heaney

, Murad

, Weaver

, Endocrine

(2011) Evaluation, treatment, and prevention of vitamin D deficiency: An Endocrine Society clinical practice guideline. J Clin Endocrinol Metab 96, 1911–1930.

53.

Pilz

, Zittermann

, Trummer

, Theiler-Schwetz

, Lerchbaum

, Keppel

, Grubler

, Marz

, Pandis

(2019) Vitamin D testing and treatment: A narrative review of current evidence. Endocr Connect 8, R27–R43.

54.

Manson

, Cook

, Lee

, Christen

, Bassuk

, Mora

, Gibson

, Gordon

, Copeland

, D’Agostino

, Friedenberg

, Ridge

, Bubes

, Giovannucci

, Willett

, Buring

, VITAL Research Group (2019) Vitamin D supplements and prevention of cancer and cardiovascular disease. N Engl J Med 380, 33–44.

55.

Marcinowska-Suchowierska

, Kupisz-Urbanska

, Luk-aszkiewicz

, Pludowski

, Jones

(2018) Vitamin D toxicity-a clinical perspective. Front Endocrinol (Lausanne) 9, 550.

56.

Crowe

, Steur

, Allen

, Appleby

, Travis

, Key

(2011) Plasma concentrations of 25-hydroxyvitamin D in meat eaters, fish eaters, vegetarians and vegans: Results from the EPIC-Oxford study. Public Health Nutr 14, 340–346.

57.

Oliai Araghi

, van Dijk

, Ham

, Brouwer-Brolsma

, Enneman

, Sohl

, Swart

, van der Zwaluw

, van Wijngaarden

, Dhonukshe-Rutten

, van Schoor

, Zillikens

, Lips

, de Groot

, Uitterlinden

, van der Velde

(2015) BMI and body fat mass is inversely associated with vitamin D levels in older individuals. J Nutr Health Aging 19, 980–985.

58.

Itkonen

, Andersen

, Bjork

, Brugard Konde

, Eneroth

, Erkkola

, Holvik

, Madar

, Meyer

, Tetens

, Torfadottir

, Thorisdottir

, Lamberg-Allardt

CJE

(2020) Vitamin D status and current policies to achieve adequate vitamin D intake in the Nordic countries. Scand J Public Health, 10.1177/1403494819896878.

59.

Brustad

, Braaten

, Lund

(2004) Predictors for cod-liver oil supplement use–the Norwegian Women and Cancer Study. Eur J Clin Nutr 58, 128–136.

60.

Geroldi

, Laakso

, DeCarli

, Beltramello

, Bianchetti

, Soininen

, Trabucchi

, Frisoni

(2000) Apolipoprotein E genotype and hippocampal asymmetry in Alzheimer’s disease: A volumetric MRI study. J Neurol Neurosurg Psychiatry 68, 93–96.

61.

Shi

, Lepore

, Gutman

, Thompson

, Baxter

, Caselli

, Wang

, Alzheimer’s Disease Neuroimaging Initiative (2014) Genetic influence of apolipoprotein E4 genotype on hippocampal morphometry: An N = 725 surface-based Alzheimer’s disease neuroimaging initiative study. Hum Brain Mapp 35, 3903–3918.

62.

Pievani

, Galluzzi

, Thompson

, Rasser

, Bonetti

, Frisoni

(2011) APOE4 is associated with greater atrophy of the hippocampal formation in Alzheimer’s disease. Neuroimage 55, 909–919.

63.

Cacciaglia

, Molinuevo

, Falcon

, Brugulat-Serrat

, Sanchez-Benavides

, Gramunt

, Esteller

, Moran

, Minguillon

, Fauria

, Gispert

, ALFA study (2018) Effects of APOE-epsilon4 allele load on brain morphology in a cohort of middle-aged healthy individuals with enriched genetic risk for Alzheimer’s disease. Alzheimers Dement 14, 902–912.