Higher prevalence of cerebral white matter hyperintensities in homozygous APOE-ɛ4 allele carriers aged 45

Abstract

Cerebral white matter hyperintensities are believed the consequence of small vessel disease and are associated with risk and progression of Alzheimer's disease. The ɛ4 allele of the APOE gene is the major factor accountable for Alzheimer's disease heritability. However, the relationship between white matter hyperintensities and APOE genotype in healthy subjects remains controversial. We investigated the association between APOE-ɛ4 and vascular risk factors with white matter hyperintensities, and explored their interactions, in a cohort of cognitively healthy adults (45–75 years). White matter hyperintensities were assessed with the Fazekas Scale from magnetic resonance images (575 participants: 74 APOE-ɛ4 homozygotes, 220 heterozygotes and 281 noncarriers) and classified into normal (Fazekas < 2) and pathological (≥2). Stepwise logistic regression was used to study the association between pathological Fazekas and APOE genotype after correcting for cardiovascular and sociodemographic factors. APOE-ɛ4 homozygotes, but not heterozygotes, bear a significantly higher risk (OR 3.432; 95% CI [1.297–9.082]; p = 0.013) of displaying pathological white matter hyperintensities. As expected, aging, hypertension and cardiovascular and dementia risk scales were also positively associated to pathological white matter hyperintensities, but these did not modulate the effect of APOE-ɛ4/ɛ4. In subjects at genetic risk of developing Alzheimer's disease, the control of modifiable risk factors of white matter hyperintensities is of particular relevance to reduce or delay dementia’s onset.

Keywords

Alzheimer’s disease Apolipoprotein E cerebrovascular magnetic resonance imaging risk factors small vessel disease

Introduction

Small vessel cerebrovascular pathology constitutes one of the most common findings in neuroimaging studies carried out in late adulthood and elderly individuals.^1,2 The most common of these lesions can be observed through computed tomography or magnetic resonance imaging (MRI) as areas of increased signal in the white matter, commonly in periventricular locations, and are referred to as white matter hyperintensities (WMH). The Fazekas scale³ rates WMH on a 0–3 point scale in the periventricular and subcortical regions combined. The Fazekas scale was designed for cross-sectional rating of WMHs, shows good intra- and inter-observer agreement and closely correlates with volumetric assessments.^4,5

It is widely accepted that the etiology of these WMH is of ischemic origin, but the precise underlying pathology is not completely understood.⁶ In this way, several alterations have been described in brain samples of deceased patients ranging from local blood brain barrier alterations with leakage of plasmatic proteins to major myelin destruction and axonal loss.^7–9 Both the prevalence and extension of WMH are strongly associated with aging and cardiovascular risk factors (CRF), particularly hypertension.¹⁰ The reported prevalence of WMH is highly variable, ranging from below 10% to over 90% of a given study’s participants. This variability can be accounted for by differences in the age of participants included in these cohorts and their comorbidities.^11,12 Additional factors could also contribute to these differences, such as ethnicity and lifestyles of the evaluated samples as well as technical differences in the MRI protocols.^13–15

From a clinical perspective, severe affectation of white matter could lead to vascular dementia and the presence of these lesions may also contribute to the development of cognitive impairment in other causes of dementia.^16,17 For instance, it has been suggested that the presence of WMH could have an additive deleterious effect to amyloid pathology in Alzheimer’s disease (AD), facilitating the progression of cognitive deterioration.^18,19 The contribution of midlife vascular risk factors in future brain function deterioration is further supported by risk scores for vascular disease like those from the Framingham study, which predict the development of cognitive deterioration.²⁰ Similarly, the CAIDE (Cardiovascular risk factors, Aging, and Incidence of DEmentia) risk score not only correlates with the extension of WMH, but also predicts the development of dementia after a follow-up of three decades.^21,22

Although the etiology of sporadic AD remains elusive, the Apolipoprotein E gene (APOE) genotype is the major known factor accountable for the heritability of the disease.²³ In this regard, the presence of one APOE-ɛ4 allele increases the risk of developing AD by a factor of 3 approximately, and homozygotes for this allele have nearly 14 times higher risk than APOE-ɛ3/ɛ3 individuals. By contrast, the APOE-ɛ2 allele seems to confer resistance towards developing the disease.²⁴ In addition to its role in sporadic AD, the APOE-ɛ4 allele has also been associated with atherosclerosis, a risk factor for the development of cerebral vascular lesions and, eventually, vascular dementia.²⁵ In this regard, it has been proposed that the APOE genotype could be an independent risk factor for the development of WMH.²⁶ Thus, the association of the APOE-ɛ4 allele with the development of AD could be explained, at least in part, by the major presence of WMH and other vascular lesions in these patients, which, in turn, could facilitate the onset of cognitive impairment and, ultimately, lead to dementia. In this way, it has been suggested that APOE-ɛ4 carriers could be more vulnerable to environmental factors that facilitate cognitive deterioration including some vascular risk factors such as fat dietary intake.²¹

Several authors have described a positive association between the APOE-ɛ4 allele and the prevalence of WMH.^27–30 However, these works frequently included patients with relevant comorbidities or advanced age that increased the prevalence of WMH and could mask the effect of the APOE genotype. Moreover, the number of subjects homozygous for the APOE-ɛ4 allele recruited in previous studies was, in general, too low to confidently evaluate whether the number of alleles could have an additive effect on WMH in the same way that occurs with the development of AD: the largest of them only included 21 homozygotes from a total of 1779 recruited subjects.²⁸

In the present work, we sought to investigate the association between the presence of the APOE-ɛ4 allele with WMH in a cohort of cognitively healthy adults aged between 45 and 75 years. In addition, we aimed to recruit a high number of APOE-ɛ4 homozygotes, with the aim of assessing whether the number of APOE-ɛ4 alleles could have an additive effect on WMH load. Finally, we aimed to study the association of vascular risk factors and their potential interaction with the APOE-ɛ4 allele.

Materials and methods

Participants

The ALFA (for ALzheimer and FAmilies) parent cohort, established by the Barcelonaβeta Brain Research Center (BBRC), consists of 2743 cognitively healthy participants, aged between 45 and 75. The ethnicity of participants is very homogeneous, and the vast majority of them are white Caucasians born in Spain and have Spanish ancestry. The ALFA study protocol was approved by the Independent Ethics Committee Parc de Salut Mar Barcelona and registered at Clinicaltrials.gov (Identifier: NCT01835717). For a complete description of the ALFA parent cohort and inclusion and exclusion criteria, please refer to Molinuevo et al.³¹ In short, participants of the present study had a Clinical Dementia Rating (CDR) equal to 0 and were cognitively normal as determined by a neuropsychological screening test battery that included the Mini-Mental State Examination, the Memory Impairment Screen, the Time Orientation of The Barcelona Test II and verbal semantic fluency. We excluded from the study those subjects with major comorbidities that could have an impact on vascular integrity. To assess whether the number of APOE-ɛ4 alleles could have an additive effect on WMH, 608 ALFA parent cohort participants without contraindications to brain MRI were selected to participate in the present study according to their APOE genotype, preferentially including APOE-ԑ4 allele carriers. The purpose of this strategy was to achieve a sample which permitted us to distinctly study the impact on WMH of APOE-ɛ4 heterozygosity and homozygosity as compared to APOE-ɛ4 allele non-carriers.

Research ethics and patient consent

The study was approved by the Ethics Committee of the “Parc de Salut Mar” (Barcelona, Spain; MRI/FBB2014v1.0) and conducted in accordance to the directives of the Spanish Law 14/2007, of 3 July, on Biomedical Research. All participants accepted the study procedures by signing an informed consent.

Cardiovascular and dementia risk factors

At ALFA baseline assessment, participants were enquired about their familiar and personal medical history and chronic medication use was recorded. Participants’ weight, height, blood pressure, and waist and hip circumference were measured. Participants were considered as hypertensive if their measured systolic blood pressure was ≥160 mmHg and/or if they were on antihypertensive medication and/or if they self-reported to suffer current hypertension. Participants were classified as having diabetes or hypercholesterolemia if they were on anti-diabetic or anti-hypercholesterolemic medication, respectively. In addition, the presence of these comorbidities was also taken as positive if self-reported. Participants’ weight and height measurements were used to calculate their body mass index (BMI). Smoking habits were collected using a simplified version of the 2009 Spanish Ministry of Health National Plan on Drugs Questionnaire. Finally, participants’ level of physical activity was measured with the Spanish short version of the Minnesota Leisure Time Physical Activity Questionnaire³² and they were categorized as active if exercised at least 150 min per week of moderate workout or 75 min per week of vigorous workout as recommended by the current guidelines.

Cardiovascular and dementia risk scores

Clinical and lifestyle-related variables gathered were used to estimate participants’ risk of suffering coronary disease events at 10 years, using the REGICOR cardiovascular risk function (an adaptation of the Framingham function validated in a Spanish sample³³), and the probability of dementia in 20 years, using the CAIDE dementia risk score, a well-established approach to assess the probability of dementia in late life according to the risk score categories in middle age.³⁴ Participants’ risk scores were estimated using both CAIDE statistical models that differ in whether they do not include (model I) or include (model II) the APOE status into consideration.

The estimated variables were as follows: (i) For the risk of suffering coronary disease events. Total cholesterol: if a participant did not self-report to suffer hypercholesterolemia and they were not on anti-hypercholesterolemic medication, the assigned cholesterol value was ≥160 and <200 mg/dl, corresponding to a coefficient of 0 in the REGICOR Cox model. If they self-reported hypercholesterolemia and/or were on anti-hypercholesterolemic medication, the assigned value was ≥240 and <280 mg/dl, corresponding to a coefficient of 0.5054 and 0.2439 for men and women, respectively, in the same model. HDL-cholesterol: Participants were assigned a ≥45 to <50 mg/dl value, corresponding to a coefficient of 0 and 0.1979 for men and women, respectively, in the REGICOR Cox model. (ii) For the probability of dementia in 20 years. Total cholesterol: if a participant did not self-report to suffer hypercholesterolemia and they were not on anti-hypercholesterolemic medication, the assigned cholesterol value was 6.5 mmol/L, corresponding to a coefficient of 0 in the CAIDE Cox model. If they self-reported hypercholesterolemia and/or were on anti-hypercholesterolemic medication, the assigned value was 6.6 mmol/L, corresponding to a coefficient of 0.631 and 0.460 in model I and II, respectively.

Apolipoprotein E genotyping

Participants’ APOE genotype was determined as described in Molinuevo et al.³¹

MRI acquisition

Scans were acquired using a 3.0-T scanner (GE Discovery MR750 W 3T). The MRI protocol was identical for all subjects and included high-resolution T1-weighted 3D structural images with an isotropic voxel size of 1 mm³ (TR/TE/TI = 8.0/3.7/450 ms, NSA = 1, Flip angle = 8° and a matrix size of 256 × 256 × 160), as well as three T2-weighted pulse sequences (256 × 256, 1 × 1 × 3 mm matrix): (i) Fluid attenuation inversion recovery (FLAIR: TR/TE/TI = 11000/90/2600 ms, Flip angle = 160°), (ii) fast spin echo (FSE: TR/TE = 5000/85 ms, Flip angle = 110°) and (iii) gradient-recalled echo (GRE: TR/TE = 1300/23 ms, Flip angle = 15°).

Radiological reporting of white matter hyperintensities

Scans were evaluated by a trained neuroradiologist within the following week after MRI acquisition and blind to the APOE-ɛ4 status of the participants. The grades of WMH were evaluated using modifications of the Fazekas Scale,³ which separately categorizes the severity of deep and periventricular lesions, on a scale from 0 to 3 (0: None or a single punctate WHM lesion, 1: Multiple punctate lesions, 2: Beginning confluency of lesions [bridging] and 3: Large confluent lesions) (Figure 1). For data analyses, the variable was dichotomized, classifying participants into two groups, namely Fazekas score <2 and score ≥2, the latter usually considered as pathological in individuals younger than 75 years old.^35–37

Figure 1.

T2-FLAIR images illustrating the Fazekas Scores (A, 0; B, 1; C, 2 and D, 3).

Statistical analyses

An initial descriptive analysis of the study sample was made and participants were divided into three groups according to the number of APOE-ɛ4 alleles of their APOE genotype. A between-group one-way ANOVA or a Kruskal–Wallis test (depending on the normality of the variables as assessed by a Kolmogorov-Smirnoff test; p < 0.05) was conducted to assess the effect of each of the variables.

A stepwise logistic regression analysis was conducted to predict the presence of a pathological Fazekas score (≥2). Participants’ APOE status was codified in three different ways that are based on (i) Model 1, the APOE haplotype (6 groups: “2/2”, “2/3”, “3/3”, “2/4”, “3/4”, “4/4”), (ii) Model 2, the number of APOE-ɛ4 alleles [3 groups: “0” (2/2, 2/3, 3/3); “1” (2/4, 3/4); “2” (4/4)] and (iii) Model 3, APOE-ɛ4 carriers/non-carriers [2 groups: “Non-carriers” (2/2, 2/3, 3/3); “Carriers” (2/4, 3/4, 4/4)]. Aside from the APOE-ɛ4 status, all models included the following predictors: hypertension, hypercholesterolemia, BMI and smoking habits, as well as sex, age and years of education as sociodemographic factors. A multinomial regression model to assess for the association of the Fazekas score as an ordinal variable was used (collapsing score = 3 with score = 2).

Regarding dementia risk estimates (CAIDE), a stepwise logistic regression analysis was also conducted, using the percentage of risk as a continuous variable. The analysis was performed for both CAIDE model I (model 4) and model II (model 5) and APOE status categorized as APOE-ԑ4/4 homozygote or not. The predictors introduced in Models 1–3 were not included here since they are already considered in the CAIDE function.

Finally, Model 6 represents a stepwise logistic regression analysis in which the estimated 10-year cardiovascular percentage of risk was used as a continuous variable. APOE status was categorized as APOE-ԑ4/4 homozygote or not. Education, the only predictor not included in the REGICOR function, was included in the analysis.

Initially, all predictors were introduced in the models along with all two-way interactions. Progressively, factors showing p values >0.1 were removed from the model except for sociodemographic factors. SPSS 17.0 for Windows was used for all the statistical analyses. Differences were considered to be statistically significant at p < 0.05 in all analyses.

Results

From the 608 ALFA parent cohort participants that were invited to take part in the present study, 575 provided valid MRI scans. The mean age was 57.79 years and 39.48% were men. Two hundred and eighty-one participants (48.9%) were APOE-ɛ4 non carriers, out of whom 7 (1.2% of the entire MRI sample) were APOE-ɛ2/2, 111 (19.3%) were APOE-ɛ2/3 and 163 (28.3%) were APOE-ɛ3/3. The group of APOE-ɛ4 carriers was composed of 294 individuals (51.1%), out of whom 174 (30.3%) were APOE-ɛ3/4, 46 (8.0%) were APOE-ɛ2/4 and 74 (12.9%) were APOE-ɛ4/4 homozygous. The main demographic characteristics of the participants are shown in Table 1. There were significant differences in age (APOE-ɛ4/4 individuals were younger than the rest), in the presence of hypercholesterolemia (the percentage of hypercholesterolemic participants increased with the number of APOE-ɛ4 alleles) and in the percentage of risk of dementia estimated by the CAIDE model II (APOE-ɛ4 allele carriers presenting a higher one). No significant between-group differences were found for the other variables. Out of the 575 subjects included in the study, 273 were rated with a Fazekas score of 0, 257 with 1, 43 with 2 and only 2 with a score of 3. Table 2 shows the distribution of Fazekas score according to number of APOE-ɛ4 alleles.

Table 1.

Characteristics of participants according to number of APOE-ɛ4 alleles.

		Number of APOE-ɛ4 alleles
	Total (n = 575)	None (n = 281)	One (n = 220)	Two (n = 74)	p
Age (years), Mean (SD)	57.79 (7.4)	58.2 (7.6)	58.3 (7.3)	54.7 (6.2)	0.001
Male sex, n (%)	227 (39.5)	102 (36.6)	98 (44.5)	27 (36.5)	0.148
Education (years), Mean (SD)	13.63 (3.5)	13.6 (3.6)	13.8 (3.5)	13.4 (3.5)	0.742
Hypertension^a, n (%)	151 (26.2)	75 (26.7)	61 (27.7)	15 (20.2)	0.440
Diabetes^b, n (%)	28 (4.9)	15 (5.3)	10 (4.5)	3 (4.1)	0.865
Cholesterol^b, n (%)	177 (30.8)	74 (26.3)	73 (33.2)	30 (40.5)	0.039
BMI, Mean (SD)	26.88 (4.1)	26.9 (3.9)	26.9 (4.3)	26.8 (4.3)	0.985
Underweight (<18.5), n (%)	1 (0.2)	0	1 (0.45)	0
Normal weight (18.5–24.9), n (%)	194 (33.7)	93 (33.1)	78 (35.4)	23 (31.1)
Overweight (25–29.9), n (%)	261 (45.4)	128 (45.5)	92 (41.8)	41 (55.4)
Obesity (≥30), n (%)	119 (20.7)	60 (21.3)	49 (22.3)	10 (13.5)
Smoking habits, n (%)					0.118
Non-smoker	92	42 (14.9)	40 (18.2)	10 (13.5)
Ex-smoker	317	161 (57.3)	118 (53.6)	38 (51.3)
Smoker	135	68 (24.2)	44 (20.0)	23 (31.1)
Not available	31	10 (3.5)	18 (8.2)	3 (4.1)
CAIDE score model I, n (%)
0–5	234 (40.7)	122 (43.4)	82 (37.3)	30 (40.5)
6–7	189 (32.9)	89 (31.7)	71 (32.3)	29 (39.2)
8–9	100 (17.4)	44 (15.6)	44 (20.0)	12 (16.2)
10–11	39 (6.8)	19 (6.6)	17 (7.7)	3 (4.1)
≥12	4 (0.69)	3 (1.2)	1 (0.4)	0
Not available	9 (1.6)	4 (1.5)	5 (2.3)	0
Risk (%), Mean (SD)	2.2 (3.2)	2.2 (2.4)	2.4 (2.4)	1.5 (1.5)	0.192
CAIDE score model II, n (%)
0–5	127 (22.1)	96 (34.2)	19 (8.6)	12 (16.2)
6–7	142 (24.7)	79 (28.1)	45 (20.4)	18 (24.3)
8–9	170 (29.6)	72 (25.6)	69 (31.4)	29 (39.2)
10–11	84 (14.6)	23 (8.2)	51 (23.4)	10 (13.5)
≥12	43 (7.5)	7 (2.5)	31 (14.1)	5 (6.7)
Not available	9 (1.6)	4 (1.4)	5 (2.3)	0
Risk (%), Mean (SD)	2.6 (3.2)	1.7 (2.0)	3.9 (4.1)	2.6 (2.3)	<0.001
REGICOR, n (%)
Low	424 (73.8)	213 (75.8)	161 (73.2)	50 (67.6)
Moderate	120 (20.8)	53 (18.9)	45 (20.4)	22 (29.7)
High	17 (2.9)	9 (3.2)	7 (3.2)	1 (1.35)
Very High	4 (0.8)	2 (0.7)	2 (0.9)	0
Not available	10 (1.7)	4 (1.4)	5 (2.27)	1 (1.35)
CV risk (%), Mean (SD)	4.3 (2.9)	4.2 (2.9)	4.5 (2.9)	4.3 (2.9)	0.428

BMI: body mass index; CV: cardiovascular.

Systolic blood pressure ≥160 mm Hg and/or use antihypertensive medication and/or self-reported.

Medications and/or self-reported.

Table 2.

Distribution of Fazekas Scale score according to number of APOE-ɛ4 alleles.

	Number of APOE-ɛ4 alleles
	None (n = 281)	One (n = 220)	Two (n = 74)
Fazekas Scale, n (%)
Score 0: None or a singlepunctate lesion	127 (45.2)	109 (49.5)	37 (50.0)
Score 1: Multiple punctate lesions	134 (47.7)	96 (43.6)	27 (36.5)
Score 2: Beginning confluency of lesions	18 (6.4)	15 (6.8)	10 (13.5)
Score 3: Large confluent lesions	2 (0.7)	0 (0.0)	0 (0.0)

Table 3 shows the results of the different binomial logistic regression models for the association of Fazekas score with APOE-ɛ4. When dichotomizing WMH as normal and pathological (Fazekas score <2 and ≥2, respectively) and coding APOE status by haplotype (Model 1), the results of the binomial logistic regression showed that only APOE-ɛ4 homozygotes (but not any other haplotype) positively correlated with the presence of pathological WMH (p = 0.011) along with age (p = 0.001) and hypertension (p = 0.007). Similarly, when coding APOE status by the number of APOE-ɛ4 alleles (Model 2), age (p = 0.001), hypertension (p = 0.008) and APOE-ɛ4/4 genotype (p = 0.008) positively correlated with the presence of pathological WMH. The results of the multinomial regression model for the association of Fazekas as an ordinal variable coding APOE status by the number of APOE-ɛ4 alleles (Supplementary Table 1) concurred with this analysis. On the other hand, when APOE-ɛ4 status was coded according to the presence of, at least, one APOE-ɛ4 allele (Model 3), no significant effect of APOE-ɛ4 was detected (p = 0.350). For all these three models, interactions did not reach the stipulated statistical threshold (p > 0.1) and, hence, were removed from the model.

Table 3.

Logistic regression models for the association of Fazekas score with APOE-ԑ4.

		B	Odds ratio (95% CI)	p
MODEL 1
Age (years)		0.083	1.087 (1.034–1.143)	0.001*
Education (years)		−0.018	0.982 (0.896–1.076)	0.693
Sex (reference: men)		−0.027	0.973 (0.501–1.890)	0.936
Hypertension		0.9198	2.506 (1.283–4.896)	0.007*
APOE-ԑ4 haplotypes	n			0.098
(reference: ɛ3/3)	163
ɛ2/2^a	7	−18.233	–	0.999
ɛ2/3	111	0.245	1.277 (0.461–3.537)	0.638
ɛ2/4	46	0.619	1.856 0.535–6.447)	0.330
ɛ3/4	174	−0.110	0.896 (0.377–2.126)	0.803
ɛ4/4	74	1.279	3.591 (1.347–9.578)	0.011*
X²= 32.227, df = 9, p < 0.001 *
MODEL 2
Age (years)		0.079	1.082 (1.032–1.136)	0.001*
Education (years)		−0.021	0.979 (0.894–1.072)	0.645
Sex (reference: men)		−0.013	0.987 (0.509–1.915)	0.970
Hypertension		0.904	2.470 (1.267–4.814)	0.008*
Number of APOE-ɛ4 alleles	n			0.015*
(reference: None)	281
One APOE-ԑ4	220	−0.022	0.978 (0.479–1.997)	0.951
Two APOE-ԑ4	74	1.191	3.289 (1.371–7.893)	0.008*
X²= 30.045, df = 6, p < 0.001 *
MODEL 3
Age (years)		0.067	1.069(1.021–1.119)	0.004*
Education (years)		−0.023	0.977 (0.893–1.069)	0.608
Sex (reference: men)		−0.020	0.980 (0.511–1.882)	0.953
Hypertension		0.883	2.418 (1.251–4.675)	0.009*
APOE-ԑ4	n
(reference: non Carriers)	281
Carriers	294	0.301	1.352(0.718–2.543)	0.350
X²= 23.579, df = 5, p < 0.001 *

Was excluded due to small n.

p ≤ 0.005.

The results of logistic regression models for the association of Fazekas score with cardiovascular and dementia risk scores and APOE- ɛ4 are shown in Table 4. As expected, homozygosity for the APOE-ɛ4 allele was positively associated with the Fazekas score [CAIDE Model I (p = 0.001), CAIDE Model II (p = 0.011) and REGICOR (p = 0.001)]. Again, interactions did not reach statistical significance and were excluded from the models.

Table 4.

Logistic regression models for the association of Fazekas score with cardiovascular and dementia risk scores and APOE-ԑ4.

		B	Odds ratio (95% CI)	p
MODEL 4
CAIDE Model I^a		0.176	1.193 (1.079–1.319)	0.001*
APOE-ԑ4	n
(reference: Others than APOE-ԑ4/4)	492
APOE-ԑ4/4	74	0.936	2.549 (1.178–5.514)	0.017*
X²= 13.979, df = 2, p = 0.001*
MODEL 5
CAIDE Model II^a		0.096	1.101 (1.022–1.186)	0.011*
APOE-ԑ4	n
(reference: Others than APOE-ԑ4/4)	492
APOE-ԑ4/4	74	0.819	2.269 (1.061–4.853)	0.035*
X²= 9.092, df = 2, p = 0.011*
MODEL 6
REGICOR^a		0.141	1.151 (1.063–1.246)	0.001*
Education		−0.027	0.973 (0.889–1.066)	0.557
APOE-ԑ4	n
(reference: Others than APOE-ԑ4/4)	492
APOE-ԑ4/4	73	0.750	2.117 (0.958–4.676)	0.064
X²= 14.634, df = 3, p = 0.002*

Percentage of cardiovascular risk.

p ≤ 0.005.

Discussion

In this work, we assessed the impact of the APOE genotype on the prevalence of pathological levels of WMH in a sample of cognitively healthy, late-middle-aged subjects and their potential interactions with vascular risk factors. To evaluate whether the number of APOE-ɛ4 alleles could have an additive effect on WMH, participants in this study were selected according to their APOE genotype, preferentially including APOE-ԑ4 allele carriers, among members of the ALFA parent cohort. This strategy resulted in a highly enriched sample for the APOE-ɛ4/4 genotype which allowed us to distinctly study the impact on WMH of APOE-ɛ4 heterozygosity and homozygosity as compared to non-carriers. Our main finding is that cognitively healthy late-middle-aged APOE-ɛ4 homozygous bear a significantly higher risk (OR 3.591; 95% Confidence Interval [1.347–9.578]; p = 0.011) of displaying pathological levels of WMH than non-carriers. On the other hand, the risk of APOE-ɛ4 heterozygous participants is not significantly different from that of non-carriers. As expected, age and hypertension were also positively associated to a pathological degree of WMH. Similarly, the observed association between presence of WMH with cardiovascular and dementia risk score estimates was also expected, in view of the existing literature.^22,38,39 Interestingly, none of the CRF analyzed here (namely diabetes, hypertension, hypercholesterolemia, BMI and smoking habits) showed an interaction with APOE status, indicating that the effect of APOE on WMH is independent of these factors.

Several studies have evaluated the association between APOE genotype and the prevalence and severity of WMH. In fact, some articles could not find any differences in WMH load between APOE-ɛ4 allele carriers and non-carriers,^40,41 but others have observed a clear correlation.^29,42 These recruited subjects from general population, patients with cerebrovascular disease⁴³ or patients affected of probable AD.⁴⁴ There is, therefore, a great variability on the characteristics of subjects evaluated among studies, including age, ethnicity, comorbidities and other demographic factors that could have an impact on WMH. Consequently, it is not surprising that previous published works have reported contradictory results regarding the impact of the APOE genotype on WMH prevalence and severity. In general, those suggesting a possible role of the APOE-ɛ4 isoform in WMH studied large cohorts recruited from the general population.³⁰ Recently, the results of a meta-analysis that addressed this topic have also reinforced the idea that, at least in general population, APOE-ɛ4 allele carriers presented a higher risk of developing WMH.²⁶ By contrast, several studies that did not find any correlations were carried out in cohorts of patients affected from cerebrovascular disease or AD. Under these conditions, other factors that participate in the development of WMH, such as cardiovascular disease, could prevail over the effect of the APOE genotype and contribute to the observed lack of association. In our study, subjects with relevant medical pathology or neurologic disease were excluded. In consequence, our sample is healthier than could be expected from an age-matched cohort selected from the general population. Moreover, participants are younger in comparison with those in previous studies. Since comorbidities and age are risk factors to develop WMH, our selection criteria produced a cohort with a lower WMH load than others reported in the literature. The relatively low level of WMH pathology in our cohort might account for the lack of correlation in heterozygosis in our results. It could be speculated that an increased risk may be found in older APOE-ɛ4 heterozygotes.

Another important consideration to take into account is the difference in the ethnicity of the subjects between studies. The association between carrying APOE-ɛ4 alleles and WMH has been clearly observed in white Caucasians, thus it cannot be a priori generalized to other populations.²⁷ In this way, other studies carried out in Japanese patients did not observe a correlation between APOE-ɛ4 and WMH.^40,45

Aside of the demographic differences between the studied cohorts, the low number of APOE-ɛ4/4 homozygotes undoubtedly represents the most relevant limitation in previous studies exploring the relationship between APOE-ɛ4 status and WMH. Very few studies have analyzed the correlation between WMH and APOE-ɛ4/4 homozygosity and the largest of them only included 21 homozygotes from a total of 1779 recruited subjects. In this work, APOE-ɛ4/4 homozygotes showed more WMH and an increased progression of the lesions during a five-year follow up period. In the same way, it was also found that WMH load of APOE-ɛ4/4 individuals is higher than that of their APOE-ɛ3/4 counterparts.²⁸ A recent report on more than 2000 participants aged between 20 and 90 years did not discover any association with the APOE genotype, purportedly because they were not able to analyse homozygotes separately (the number of homozygotes is not reported in the paper).⁴¹ Therefore, the major strength of our study is the number of recruited APOE-ɛ4/4 homozygous (n = 74) which is, to our knowledge, considerably higher than any other single center study published to date. This enabled us to characterize the association between APOE-ɛ4 genetic load and WMH. In fact, no correlation with pathological WMH could be detected in our sample in association with APOE-ɛ4 when we pooled all subjects presenting at least one ɛ4 allele together. Nevertheless, it is worth mentioning that our sampling strategy was designed to strengthen the ability to detect associations for APOE-ɛ4 homozygotes. The limitation of this strategy is that the allele frequencies in our study are not representative of an unselected population, thus increasing the risk of bias and preventing us from drawing any epidemiological conclusions. In addition, even though the vast majority of the study participants are white Caucasians and, consequently, ethnically homogeneous, specific genotyping of ancestry informative markers in our study population has not been performed. Given that APOE alleles are highly stratified by ethnicity⁴⁶ even within white European populations, unmeasured social, cultural or genetic factors that vary by ethnicity and increase WMH could confound the observed associations.

On the other hand, a few studies have suggested an association between the APOE-ɛ2 allele and increased levels of WMH as well as with other vascular cerebral alterations.⁴⁷ Our observations do not support this hypothesis since no correlation was found between WHM and this allele. However, other vascular lesions such as microbleeds or lacunar infarcts were too scarce in our cohort to allow us to perform correlational analyses.

The increased WMH observed in APOE-ɛ4/4 individuals could be related to an increased tendency towards vascular pathology, a higher sensibility to risk factors, a reduced resilience to brain injury and a reduced recovery of lesions.⁴⁸ In any case, the accumulation of WMH could decrease the efficiency of subcortical white matter to perform its function and may have additive or even synergistic effects with other noxious processes acting in the brain like neurodegeneration in the cortical grey matter.¹⁶ In the last instance, accumulation of WMH could facilitate cognitive deterioration and the onset of AD dementia.^49,50 Hence, controlling WMH load in AD susceptible individuals could be a useful preventive strategy to reduce or delay cognitive decline.⁵¹ In contrast with age or APOE genotype, other factors associated with WMH, like hypertension, are modifiable with changes in life style or pharmacological treatment. There is no doubt that the control of vascular risk factors is the way to go for all patients not only for these with increased risk of developing WMH. However, it could be theorized that a more strict control of these factors could particularly benefit the APOE-ɛ4/4 homozygotes. The same rationale could be considered for the development of treatments focused on reducing WMH load and progression. Such interventions could reduce the likelihood of cognitive deterioration and eventually the onset of dementia in APOE-ɛ4/4 homozygotes, which are at a higher risk of developing AD.²⁸

Cognitively healthy APOE-ɛ4 homozygotes have also been reported to show a significantly higher prevalence of cerebral amyloid pathology. At the age of 55, approximately 50% of APOE-ɛ4/ɛ4 individuals display abnormal levels of amyloid biomarkers, as compared to only 10% of APOE-ɛ3/ɛ3 persons and about 20% of carriers of a single APOE-ɛ4 allele.⁵² Cerebral amyloid pathology has also been pinpointed as a relevant factor in the development of WMH in cognitively normal elderly individuals.^53,54 In this regard, the main limitation of our work is that we could not assess the impact in the prevalence of WMH of cerebral amyloid pathology independently of that of APOE-ɛ4 status, hypertension and age.

Conclusions

Cognitively healthy, late-middle-aged APOE-ɛ4 homozygotes show a higher risk of presenting pathological levels of WMH as compared to their heterozygous and non-carrier counterparts. This association was found after controlling for well-established risk factors for the development of WMH such as age and hypertension and was also independent from other CRF. Given the known association between WMH and future cognitive decline, the control of modifiable risk factors in individuals at higher risk of developing WMH appears as a useful preventive strategy to reduce or delay the onset of dementia.

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research leading to these results has received funding from “la Caixa” Foundation. Additional funding was obtained from Fondo de Investigación Sanitaria (FIS), Instituto de Salud Carlos III (ISC-III) under grant PI12/00326 and Barcelona city council under agreement #0724/13. Juan D Gispert holds a ‘Ramón y Cajal’ fellowship (grant no. RYC-2013-13054).

Acknowledgements

This publication is part of the ALFA study (ALzheimer and FAmilies). The authors would like to express their most sincere gratitude to the ALFA project volunteers, without whom this research would have not been possible.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Authors’ contributions

SR and AB-R contributed equally to this work. All authors listed (SR, AB-R, NB, CM, AT, CF, AC, SM, ME, NG, KF, JC, JLM and JDG) made a substantial contribution to the concept and design, acquisition of data or analysis and interpretation of data, drafted the article or revised it critically for important intellectual content and approved the version to be published.

Supplementary material

Supplementary material for this paper can be found at the journal website:

References

Sandeman

Hernandez Mdel

Morris

et al.

Incidental findings on brain MR imaging in older community-dwelling subjects are common but serious medical consequences are rare: a cohort study. PLoS One 2013; 8: e71467.

Vernooij

Ikram

Tanghe

et al.

Incidental findings on brain MRI in the general population. N Engl J Med 2007; 357: 1821–1828.

Fazekas

Chawluk

Alavi

et al.

MR signal abnormalities at 1.5 T in Alzheimer's dementia and normal aging. AJR Am J Roentgenol 1987; 149: 351–356.

Prins

Scheltens

. White matter hyperintensities, cognitive impairment and dementia: an update. Nat Rev Neurol 2015; 11: 157–165.

Kapeller

Barber

Vermeulen

et al.

Visual rating of age-related white matter changes on magnetic resonance imaging: scale comparison, interrater agreement, and correlations with quantitative measurements. Stroke 2003; 34: 441–445.

Bernbaum

Menon

Fick

et al.

Reduced blood flow in normal white matter predicts development of leukoaraiosis. J Cereb Blood Flow Metab 2015; 35: 1610–1615.

Huisa

Caprihan

Thompson

et al.

Long-term blood-brain barrier permeability changes in Binswanger disease. Stroke 2015; 46: 2413–2418.

Haller

Kovari

Herrmann

et al.

Do brain T2/FLAIR white matter hyperintensities correspond to myelin loss in normal aging? A radiologic-neuropathologic correlation study. Acta Neuropathol Commun 2013; 1: 14.

Murray

Vemuri

Preboske

et al.

A quantitative postmortem MRI design sensitive to white matter hyperintensity differences and their relationship with underlying pathology. J Neuropathol Exp Neurol 2012; 71: 1113–1122.

10.

Liao

Cooper

Cai

et al.

Presence and severity of cerebral white matter lesions and hypertension, its treatment, and its control. The ARIC Study. Atherosclerosis Risk in Communities Study. Stroke 1996; 27: 2262–2270.

11.

de Leeuw

de Groot

Achten

et al.

Prevalence of cerebral white matter lesions in elderly people: a population based magnetic resonance imaging study. The Rotterdam Scan Study. J Neurol Neurosurg Psychiatr 2001; 70: 9–14.

12.

Hopkins

Beck

Burnett

et al.

Prevalence of white matter hyperintensities in a young healthy population. J Neuroimag 2006; 16: 243–251.

13.

Brickman

Schupf

Manly

et al.

Brain morphology in older African Americans, Caribbean Hispanics, and whites from northern Manhattan. Arch Neurol 2008; 65: 1053–1061.

14.

Gardener

Scarmeas

et al.

Mediterranean diet and white matter hyperintensity volume in the Northern Manhattan Study. Arch Neurol 2012; 69: 251–256.

15.

Tian

Glynn

Erickson

et al.

Objective measures of physical activity, white matter integrity and cognitive status in adults over age 80. Behav Brain Res 2015; 284: 51–57.

16.

Debette

Beiser

DeCarli

et al.

Association of MRI markers of vascular brain injury with incident stroke, mild cognitive impairment, dementia, and mortality: the Framingham Offspring Study. Stroke 2010; 41: 600–606.

17.

Inzitari

Pracucci

Poggesi

et al.

Changes in white matter as determinant of global functional decline in older independent outpatients: three year follow-up of LADIS (leukoaraiosis and disability) study cohort. BMJ 2009; 339: b2477.

18.

Tosto

Zimmerman

Hamilton

et al.

The effect of white matter hyperintensities on neurodegeneration in mild cognitive impairment. Alzheimers Dement 2015; 11: 1510–1519.

19.

Brickman

Zahodne

Guzman

et al.

Reconsidering harbingers of dementia: progression of parietal lobe white matter hyperintensities predicts Alzheimer's disease incidence. Neurobiol Aging 2015; 36: 27–32.

20.

Kaffashian

Dugravot

Elbaz

et al.

Predicting cognitive decline: a dementia risk score vs. the Framingham vascular risk scores. Neurology 2013; 80: 1300–1306.

21.

Kivipelto

Rovio

Ngandu

et al.

Apolipoprotein E epsilon4 magnifies lifestyle risks for dementia: a population-based study. J Cell Mol Med 2008; 12: 2762–2771.

22.

Vuorinen

Spulber

Damangir

et al.

Midlife CAIDE dementia risk score and dementia-related brain changes up to 30 years later on magnetic resonance imaging. J Alzheimers Dis 2015; 44: 93–101.

23.

Corder

Saunders

Strittmatter

et al.

Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer's disease in late onset families. Science 1993; 261: 921–923.

24.

Farrer

Cupples

Haines

et al.

Effects of age, sex, and ethnicity on the association between apolipoprotein E genotype and Alzheimer disease. A meta-analysis. APOE and Alzheimer Disease Meta Analysis Consortium. JAMA 1997; 278: 1349–1356.

25.

Karvonen

Kauma

Kervinen

et al.

Apolipoprotein E polymorphism affects carotid artery atherosclerosis in smoking hypertensive men. J Hypertens 2002; 20: 2371–2378.

26.

Schilling

DeStefano

Sachdev

et al.

APOE genotype and MRI markers of cerebrovascular disease: systematic review and meta-analysis. Neurology 2013; 81: 292–300.

27.

Hogh

Garde

Mortensen

Jorgensen

et al.

The apolipoprotein E epsilon4-allele and antihypertensive treatment are associated with increased risk of cerebral MRI white matter hyperintensities. Acta Neurol Scand 2007; 115: 248–253.

28.

Godin

Tzourio

Maillard

et al.

Apolipoprotein E genotype is related to progression of white matter lesion load. Stroke 2009; 40: 3186–3190.

29.

Brickman

Schupf

Manly

et al.

APOE epsilon4 and risk for Alzheimer's disease: do regionally distributed white matter hyperintensities play a role?

Alzheimers Dement 2014; 10: 619–629.

30.

de Leeuw

Richard

de Groot

et al.

Interaction between hypertension, apoE, and cerebral white matter lesions. Stroke 2004; 35: 1057–1060.

31.

Molinuevo

Gramunt

Gispert

et al.

The ALFA project: a research platform to identify early pathophysiological features of Alzheimer's disease. Alzheimer’s & Dementia: Translational Research & Clinical Interventions 2016; 2: 82–92.

32.

Elosua

Garcia

Aguilar

et al.

Validation of the Minnesota Leisure Time Physical Activity Questionnaire In Spanish Women. Investigators of the MARATDON Group. Med Sci Sports Exerc 2000; 32: 1431–1437.

33.

Marrugat

Vila

Baena-Diez

et al.

[Relative validity of the 10-year cardiovascular risk estimate in a population cohort of the REGICOR study]. Rev Esp Cardiol 2011; 64: 385–394.

34.

Kivipelto

Ngandu

Laatikainen

et al.

Risk score for the prediction of dementia risk in 20 years among middle aged people: a longitudinal, population-based study. Lancet Neurol 2006; 5: 735–741.

35.

Schmidt

Enzinger

Ropele

et al.

Progression of cerebral white matter lesions: 6-year results of the Austrian Stroke Prevention Study. Lancet 2003; 361: 2046–2048.

36.

Schmidt

Petrovic

Ropele

et al.

Progression of leukoaraiosis and cognition. Stroke Sep; 38: 2619–2625.

37.

Schmidt

Haybaeck

et al.

Heterogeneity in age-related white matter changes. Acta Neuropathol 2011; 122: 171–185.

38.

Jeerakathil

Wolf

Beiser

et al.

Stroke risk profile predicts white matter hyperintensity volume: the Framingham Study. Stroke 2004; 35: 1857–1861.

39.

Homayoon

Ropele

Hofer

et al.

Microstructural tissue damage in normal appearing brain tissue accumulates with Framingham Stroke Risk Profile Score: magnetization transfer imaging results of the Austrian Stroke Prevention Study. Clin Neurol Neurosurg 2013; 115: 1317–1321.

40.

Hirono

Yasuda

Tanimukai

et al.

Effect of the apolipoprotein E epsilon4 allele on white matter hyperintensities in dementia. Stroke 2000; 31: 1263–1268.

41.

Habes

Erus

Toledo

et al.

White matter hyperintensities and imaging patterns of brain ageing in the general population. Brain 2016; 139: 1164–1179.

42.

Lyall

Munoz Maniega

Harris

et al.

APOE/TOMM40 genetic loci, white matter hyperintensities, and cerebral microbleeds. Int J Stroke 2015; 10: 1297–1300.

43.

Wen

Baum

Cheung

et al.

Apolipoprotein E epsilon4 allele is associated with the volume of white matter changes in patients with lacunar infarcts. Eur J Neurol 2006; 13: 1216–1220.

44.

Morgen

Schneider

Frolich

et al.

Apolipoprotein E-dependent load of white matter hyperintensities in Alzheimer's disease: a voxel-based lesion mapping study. Alzheimers Res Ther 2015; 7: 27.

45.

Sawada

Udaka

Izumi

et al.

Cerebral white matter lesions are not associated with apoE genotype but with age and female sex in Alzheimer's disease. J Neurol Neurosurg Psychiatr 2000; 68: 653–656.

46.

Biffi

Sonni

Anderson

et al.

Variants at APOE influence risk of deep and lobar intracerebral hemorrhage. Ann Neurol 2010; 68: 934–943.

47.

Lemmens

Görner

Schrooten

et al.

Association of Apolipoprotein E ɛ2 With White Matter Disease but Not With Microbleeds. Stroke 2007; 38: 1185–1188.

48.

DeCarli

Reed

Miller

et al.

Impact of apolipoprotein E epsilon4 and vascular disease on brain morphology in men from the NHLBI twin study. Stroke 1999; 30: 1548–1553.

49.

Dufouil

Godin

Chalmers

et al.

Severe cerebral white matter hyperintensities predict severe cognitive decline in patients with cerebrovascular disease history. Stroke 2009; 40: 2219–2221.

50.

Staekenborg

Koedam

Henneman

et al.

Progression of mild cognitive impairment to dementia: contribution of cerebrovascular disease compared with medial temporal lobe atrophy. Stroke 2009; 40: 1269–1274.

51.

Provenzano

Muraskin

Tosto

et al.

White matter hyperintensities and cerebral amyloidosis: necessary and sufficient for clinical expression of Alzheimer disease?

JAMA Neurol 2013; 70: 455–461.

52.

Jansen

Ossenkoppele

Knol

et al.

Prevalence of cerebral amyloid pathology in persons without dementia: a meta-analysis. JAMA 2015; 313: 1924–1938.

53.

Marnane

Al-Jawadi

Mortazavi

et al.

Periventricular hyperintensities are associated with elevated cerebral amyloid. Neurology 2016; 86: 535–543.

54.

Scott

Braskie

Tosun

et al.

Cerebral amyloid and hypertension are independently associated with white matter lesions in elderly. Front Aging Neurosci 2015; 7: 221.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.25 MB

Higher prevalence of cerebral white matter hyperintensities in homozygous APOE-ɛ4 allele carriers aged 45–75: Results from the ALFA study

Abstract

Keywords

Introduction

Materials and methods

Participants

Research ethics and patient consent

Cardiovascular and dementia risk factors

Cardiovascular and dementia risk scores

Apolipoprotein E genotyping

MRI acquisition

Radiological reporting of white matter hyperintensities

Statistical analyses

Results

Discussion

Conclusions

Footnotes

Funding

Acknowledgements

Declaration of conflicting interests

Authors’ contributions

Supplementary material

References

Supplementary Material