Apolipoprotein E Gene Polymorphisms Are Associated With Carotid Plaque Formation but Not With Intima-Media Wall Thickening
Results From the Perth Carotid Ultrasound Disease Assessment Study (CUDAS)
Background and Purpose— Several studies have investigated the role of apolipoprotein E (apoE) polymorphisms on carotid intima-media thickness (IMT) with conflicting results. The objective of this study was to use a large, community-based population to investigate associations between apoE gene polymorphisms and cardiovascular disease–associated phenotypes: IMT, carotid artery plaque, and low- (LDL-C) and high-density lipoprotein cholesterol (HDL-C).
Methods— ApoE genotypes were determined in 1109 randomly selected community subjects with an equal man-to-woman ratio and equal numbers in each age decile who were 27 to 77 years of age and had bilateral carotid B-mode ultrasound and cardiovascular risk factor measurements.
Results— Multivariate analyses, stratified by sex, demonstrated an association between apoE genotypes and LDL-C levels in men (P=0.03) and women (P<0.001). A significant linear trend in increasing LDL-C (β=0.33 per unit change in genotype; SE=0.07; P<0.001) levels with increasing number of ε4 alleles across the ε3/ε3, ε3/ε4, or ε4/ε4 genotypes was observed in women but not in men. The associations were independent of age, diastolic blood pressure, and history of diabetes mellitus. Multivariate analyses found a log-additive trend in risk of developing carotid plaque with increasing numbers of ε4 alleles across the ε3/ε3, ε3/ε4, and ε4/ε4 genotypes (odds ratio [OR], 1.72 per unit change in genotype; 95% CI, 1.05 to 2.80; P=0.03) in men. There was no association between plaque frequency and the ε4 allele in women. However, the ε2/ε3 genotype was shown to be associated with a lower OR (OR, 0.40; 95% CI, 0.17 to 0.91; P=0.03) for carotid plaques relative to the ε3/ε3 genotype in women. The associations were independent of age and standard vascular risk factors. There were no significant independent associations between apoE genotypes and IMT in either men or women.
Conclusions— Our data suggest that polymorphisms in the apoE gene are significantly associated with LDL-C levels and increased risk of carotid plaque formation in men but not IMT in either men or women.
Apolipoprotein E (apoE) plays an important role in the development of atherosclerosis.1 It aids in cholesterol homeostasis2,3 and is a component of chylomicrons, very low density lipoprotein (VLDL), and their degradation products, high- (HDL-C) and low-density lipoprotein cholesterol (LDL-C). Several hepatic receptors recognize apoE, including the LDL receptor, LDL receptor–related protein, and the VLDL receptor, where it acts as a ligand for receptor-mediated clearance of lipoproteins.
ApoE is synthesized endogenously in foam cells and after stimulation by extracellular lipid-free apoA-1. This facilitates cholesterol efflux from lipid-laden foam cells, within the intima of lesion, into circulation via HDL-containing apoA-1.4 Exogenous apoE can assist in cholesterol transport from lesional foam cells as an extracellular, free cholesterol acceptor.
ApoE also directly modifies macrophage and T-lymphocyte–mediated immune responses to inflammatory atherosclerosis. The production of apoE in macrophages is regulated by inflammatory cytokines, interferon-γ, and tumor necrosis factor-α. ApoE can also inhibit the proliferation of CD4 and CD8 T lymphocytes in lesions.4 ApoE has antioxidant activity demonstrated in a dose-dependent manner, with the E2 isoform having the greatest activity and E4 having the least.5
The apoE gene encodes 3 alleles: ε2, ε3, and ε4. The ε2 allele is associated with lower LDL-C levels, and the ε4 allele is associated with higher levels relative to the ε3 allele.3 This observation led to the hypothesis that apoE may play an important role in the development of coronary heart disease. It has been estimated that carriers of the ε4 allele have a 1.4-times–higher risk of developing coronary heart disease than ε3 carriers.6
Most studies to date investigating the role of apoE genotypes in the development of atherosclerosis7–17 have used small numbers of subjects with differing conclusions. In this study, a population (Perth Carotid Ultrasound Disease Assessment Study [CUDAS] population18) of 1111 randomly selected male and female community subjects 27 to 77 years of age was used to investigate the association of apoE polymorphisms with carotid intima-media thickness (IMT) and the presence of carotid plaques. All subjects had high-resolution bilateral B-mode carotid ultrasound examination and a comprehensive risk factor and biochemical assessment and were genotyped for apoE.
CUDAS consisted of 1111 subjects at an equal man-to-woman ratio and equal numbers in each age decile between 20 and 70 years.18 The study protocol was approved by the Institutional Ethics Committee of the University of Western Australia, and written, informed consent was obtained from all study participants. A self-administered questionnaire, anthropomorphic measurements, and the lower of 2 resting blood pressures were used as previously described.18,22
Samples for total cholesterol (TC), HDL-C, triglycerides, LDL-C, and homocysteine were collected and analyzed as previously described.18 Genomic DNA from 1109 subjects was genotyped as previously described.20
Bilateral carotid B-mode ultrasound was performed by 2 trained sonographers using a 7.5-MHz annular phased-array transducer on an Interspec (Apogee) CX 200 ultrasound machine as previously described.18 Plaque was defined as a clearly identified area of focal increased thickness (>1 mm) of the intima-media layer. IMT was defined as the distance between the characteristic echoes from the lumen-intima and media-adventitia interfaces.18 The intraobserver coefficients of variability were 2.9% for sonographer 1 and 4.8% for sonographer 2. The interobserver coefficient of variability was 5.9%.
The primary continuous outcome variables of the association analyses were mean IMT, LDL-C, HDL-C, and TC levels. The primary binary outcome variables were the presence or absence of carotid plaque. ApoE genotype, the principal explanatory variable, was coded into 6 classes (ε3/ε3=0, ε2/ε2=1, ε2/ε3=2, ε2/ε4=3, ε3/ε4=4, ε4/ε4=5) and analyzed categorically as a binary dummy variable relative to the most common ε3/ε3 genotype. In secondary analyses for linear or additive trends for the ε3 and ε4 alleles, apoE genotypes were analyzed as a continuous covariate (ε3/ε3=0, ε3/ε4=1, ε4/ε4=2).
Sex, carotid plaque, physician-diagnosed diabetes, and hypertension were analyzed as binary variables. All other variables were analyzed as continuous. Cigarette smoking status was analyzed as both a binary variable (smoking ever) and a continuous variable (pack-years).
Bivariate analysis used analysis of variance to compare mean IMT across all apoE genotypes and either χ2 tests or Fisher’s exact test on contingency tables to compare genotypes with categorical variables. Initial bivariate analyses were also stratified by sex.
Hardy-Weinberg equilibrium was tested by use of a χ2 goodness-of-fit test on a contingency table of observed versus predicted genotype frequencies.
Generalized linear models (linear and logistic regression)19 were used to model the effects of multiple covariates on the continuous and dichotomous outcomes. Multivariate risk predictors of IMT and plaque by logistic regression have previously been reported.18,20,21 Both forward and backward stepwise modeling procedures were used to select a subset of independent predictors of an outcome of interest. Checks of goodness of fit22 included an investigation of the need for interaction or polynomial terms, analyses of Pearson residuals, and examination of the effect of observations with high regression leverage. Potential interactions of measured genotypes with other covariates (eg, age, sex, LDL-C level, cigarette smoking) were explicitly investigated by the inclusion of relevant interaction terms in the models. The Generalized linear models were also used to estimate genotype-specific means or odds ratios (ORs).
Minitab for Windows, version 12.1 (Minitab Inc), and S-Plus v4.5 (Mathsoft Inc) were used to manage and analyze data. Statistical significance was defined at the standard 5% level.
The study population characteristics are described in Table 1. The mean age was 53.3 years (SD, 12.7 years), with nearly equal numbers of subjects in each age decile. The sex ratio was balanced with 558 male subjects (50.3%), and 25.6% of the population had ≥1 detectable carotid plaques.
The allele ε2 made up 8.0%, ε3 composed 77.3%, and ε4 made up 14.7% of the total alleles. The apoE genotype frequencies were consistent with Hardy-Weinberg equilibrium (χ25=3.30, P=0.65) and not significantly different in men and women (χ25=1.45, P=0.92). The ε2/ε2 genotype was excluded from analyses because of its very low frequency (Table 2).
Multivariate modeling suggested that increasing age, male sex, increasing diastolic blood pressure, and diabetes were important independent predictors of increased TC, LDL-C, and HDL-C levels. TC, LDL-C, and HDL-C levels also increased more rapidly with age in women than in men (data not shown).
Bivariate analysis stratified by sex suggested an association between the apoE genotypes and TC in women, but an association with LDL-C was present in both men and women (Table 2). Multivariate analyses stratified by sex confirmed this association with LDL-C in men and women and TC in women (Table 3). Secondary analyses in women with ε3/ε3, ε3/ε4, or ε4/ε4 genotypes demonstrated a significant increasing linear trend with TC (β=0.37 per unit change in genotype; SE=0.08; P<0.001) and LDL-C (β=0.33 per unit change in genotype; SE=0.07; P<0.001). No linear trend for LDL-C or TC was found in men (Tables 2 and 3). There was no significant difference in triglyceride levels with apoE genotypes in either men (F4,549=1.33, P=0.26) or women (F4,544=0.07, P=0.99).
Carotid Artery Plaques
Multivariate modeling had previously suggested21 that age, LDL-C, pack-years of cigarette smoking, hypertension, systolic blood pressure, diabetes, and history of vascular disease were important independent predictors of carotid plaques in this population.
Bivariate analysis indicated that apoE genotypes were not significantly associated with carotid plaque in the whole population (χ24=6.70, P=0.15). Furthermore, when stratified by sex, there was no significant association of apoE genotypes with carotid plaque in either men or women (Table 2).
Multivariate analyses stratified by sex and adjusted for important covariates (Table 4) gave some evidence of significant, but differing, associations of apoE genotypes and carotid plaque in both men and women. In men, the ε3/ε4 heterozygote was associated with a 1.79-fold–increased (95% CI, 1.01 to 3.17; P=0.05) risk of carotid plaque relative to the ε3/ε3 homozygote (Table 4). A secondary analysis in men with ε3/ε3, ε3/ε4, or ε4/ε4 genotypes showed a significant linear trend with increasing risk of carotid plaque (OR, 1.72 per unit change in genotype; 95% CI, 1.05 to 2.80; P=0.03). In contrast, there was no significant effect of the ε4 allele on the presence of carotid plaque in women (Table 4). However, multivariate analysis indicated the ε2/ε3 genotype was associated with significantly decreased risk of carotid plaque relative to the ε3/ε3 genotype in women (OR, 0.40; 95% CI, 0.17 to 0.91; P=0.03) (Table 4).
Inclusion of a sex–apoE genotype interaction term (genotype coded as a continuous covariate) in a model including men and women as a formal test of interaction indicated that the interaction of apoE genotype with sex was significant for risk of carotid plaque (P=0.03). After adjustment for all relevant covariates (including sex and apoE genotypes as main effects), the multiplicative increase in carotid plaque risk was ≈1.24 times greater per unit change in genotype in women than in men (95% CI, 1.02 to 1.50).
Removal of LDL-C levels as a covariate term in the multivariate models given in Table 4 gave essentially unchanged results. The apoE genotypes showed no significant evidence of interaction with other covariates independently associated with carotid plaque in the multivariate generalized linear models (data not shown).
Mean Carotid IMT
Bivariate analysis indicated that apoE genotypes were marginally associated with mean IMT in the whole population (F4,1099=2.74, P=0.03); ε4/ε4 was associated with significantly higher IMT than other genotypes. Further bivariate analysis stratified by sex suggested that this association was not present in men (P=0.07) or in women (P=0.58) (Table 2). Multivariate analyses also found no significant association between apoE genotypes and mean IMT after adjustment for the important covariates in the whole population (F4,1086=0.70, P=0.60), in men (F4,539=0.61, P=0.66), or in women (F4,534=0.30, P=0.88). Removal of LDL-C and HDL-C levels as covariate terms from the multivariate model did not substantially change these results. ApoE genotypes showed no significant evidence of interaction with any of the other covariates independently associated with carotid IMT (data not shown).
This study of a large, randomly selected community population has demonstrated an independent association of apoE genotypes with LDL-C levels and the presence of plaque but not carotid IMT.
The allele and genotype frequencies in our sample were similar to those previously reported.3,23 The frequency of the ε2/ε2 genotype in our population, as in others, was very low (0.5%) and was excluded from all analyses.
Bivariate analysis indicated apoE genotypes were associated with TC and LDL-C but not with HDL-C in the whole population. There have been mixed results from studies investigating variability in HDL-C levels with apoE genotypes. Some studies have found an association with HDL-C levels and apoE genotypes,24,25 whereas others have found no association.26–28
Multivariate analyses stratified by sex confirmed the association of apoE genotypes and increased LDL-C levels in men and women and with increased TC in women (Table 3). Secondary analyses found some evidence of a linear trend in increasing TC and LDL-C levels with an increasing number of ε4 alleles in women. Generally, most other studies have found that the ε4 allele was associated with higher LDL-C levels and/or that the ε2 allele was associated with lower levels relative to the ε3 allele.2,3,7,8,26 Most of these studies either ignored the ε4/ε4 genotype or grouped it with the ε3/ε4 genotype. In this study, grouping the genotypes in this way did not significantly change the results (data not shown).
A comparison of TC levels with apoE genotypes in 9 populations25 found a significant decrease in TC levels with the ε2 allele and an increase with the ε4 allele relative to the ε3/ε3 genotype. This trend was similar in our population, but when stratified by sex, this trend was seen only in women. Our results agree with the Framingham Offspring Study23 in which apoE genotypes were associated with TC in women only and with LDL-C in men and women.
In men, there was a log-additive trend in the risk of carotid plaque with increasing numbers of ε4 alleles. In women, the ε2/ε3 genotype was associated with a lower risk for plaque development relative to the ε3/ε3 genotype (Table 4). These two associations were independent of the effects of the apoE genotypes on increasing LDL-C levels. These associations were apparent only in multivariate analysis, suggesting them to be weak. The sex-dependent effects of apoE genotypes on the presence of carotid plaque could be explained by the effects of sex hormones on gene expression. Estradiol has been shown to upregulate apoE gene expression in mouse models.29
Two other studies have investigated the effect of apoE genotypes on the presence of plaque. One found no association between apoE genotypes and plaque.14 This small study of 133 men and 92 women lacked sufficient power to detect an association. A large study of 5401 men and women (mean age, 69.2 years)17 demonstrated that the risk of developing >3 plaques was decreased in carriers of ε2/ε3 (OR, 0.6; 95% CI, 0.4 to 0.8; P=0.003) for the whole population but not when men and women were analyzed separately.
Multivariate analyses showed that apoE genotypes were not associated with carotid IMT in the whole population or when stratified by sex. LDL-C is an independent predictor of IMT, and apoE genotypes are an independent predictor of LDL-C levels; however, apoE genotypes did not affect LDL-C levels enough to produce a significant change in IMT. It is of interest that the apoE genotypes were associated with the presence of plaque but not with increased IMT. This may be explained by the apoE genotypes being expressed in lesional foam cell macrophages30 but not in normal arterial intima31 and may play a role only in plaque formation.
Eleven other studies have investigated apoE genotypes and carotid IMT with varying results. These studies reached 4 different conclusions: the ε4 allele was associated with increased IMT6,8,9 ; the ε2 allele was associated with increased IMT10,11 ; the ε2 allele was associated with decreased IMT6,17 ; and no association between IMT and apoE genotype.12–14 Another study16 found that the ε3 allele was weakly associated with increased right but not left carotid IMT. The inconsistent results from these studies are most likely due to the differences in study groups—eg, in age, sex, race, and IMT measurement criteria—and small populations, with 1 exception.17
This study demonstrates that LDL-C was significantly associated with the apoE genotypes in both women and men, with a significant increasing linear trend across the ε3/ε3, ε3/ε4, and ε4/ε4 genotypes in women. ApoE genotypes were not independently associated with carotid IMT. In men, the ε3/ε4 genotype was associated with an increased risk of carotid plaque relative to the ε3/ε3 genotype; in women, however, the ε2/ε3 genotype was associated with a significant decrease in plaque frequency. These latter associations were independent of age, smoking history, blood pressure, LDL-C, or diabetes. Our results therefore suggest that the effects of apoE alleles on the cardiovascular outcomes investigated are not mediated through these related factors. This information may assist in disentangling the complex pathogenetic pathways involved in hyperlipidemia and carotid plaque formation.
This study was supported by grants in aid from the National Heart Foundation of Australia (G94P 4232, G97P 5002) and Healthway (J.H., P.L.T., J.P.B.). We thank Jo Crittenden and Marcus Sommerville for their invaluable technical assistance.
- Received April 5, 2002.
- Revision received October 15, 2002.
- Accepted October 29, 2002.
Davignon J, Gregg RE, Sing SF. Apolipoprotein E polymorphism and atherosclerosis. Arteriosclerosis. 1988; 8: 1–21.
Terry JG, Howard G, Mercuri M, Bond MG, Crouse JR. Apolipoprotein E polymorphism is associated with segment-specific extracranial carotid artery intima-media thickening. Stroke. 1996; 27: 1755–1759.
Cattin L, Fisicaro M, Tonizzo M, Valenti M, Danek GM, Fonda M, Da Col PG, Casagrande S, Pincetti E, Bovenzi M, Baralle F. Polymorphism of the apolipoprotein E gene and early carotid atherosclerosis defined by ultrasonography in asymptomatic adults. Arterioscler Thromb Vasc Biol. 1997; 17: 91–94.
Kogawa K, Nishizawa Y, Hosoi M, Kawagisha T, Maekawa K, Shoji T, Okuno Y, Morii H. Effect of polymorphism of apolipoprotein E and angiotensin-converting enzyme genes on arterial wall thickness. Diabetes. 1997; 46: 682–687.
Sass C, Zannad F, Herbeth B, Salah D, Chapet O, Siest G, Visvikis S. Apolipoprotein E4, lipoprotein lipase C447 and angiotensin-I converting enzyme deletion alleles were not associated with increased wall thickness of carotid and femoral arteries in healthy subjects from the Stanislas cohort. Atherosclerosis. 1998; 140: 89–95.
Slooter AJC, Bots ML, Havekes LM, A.Iglesias del Sol, Cruts M, Grobbee DE, Hofman A, MD, Van Broeckhoven C, Witteman JCM, van Duijn C. Apolipoprotein E and carotid artery atherosclerosis: the Rotterdam Study. Stroke. 2001; 32: 1947–1952.
McQuillan BM, Beilby JP, Nidorf M, Thompson PL, Hung J. Hyperhomocysteinemia but not the C677T mutation of methylenetetrahydrofolate reductase is an independent risk determinant of carotid wall thickening: the Perth Carotid Ultrasound Disease Assessment Study (CUDAS). Circulation. 1999; 99: 2383–2388.
Armitage P, Barry G. Statistical Methods in Medical Research. Oxford, UK: Blackwell Scientific Publications; 1994.
Hung J, McQuillan BM, Nidorf M, Thompson PL, Beilby JP. Angiotensin-converting enzyme gene polymorphism and carotid wall thickening in a community population. Arterioscler Thromb Vasc Biol. 1999; 340: 14–22.
McCullagh P, Nelder J. Generalised Linear Models. London, UK: Chapman and Hall; 1989.
Schaefer EJ, Lamon-Fava S, Johnson S, Ordovas JM, Schaefer MM, Castelli WP, Wilson WF. Effects of gender and menopausal status on the association of apolipoprotein E phenotype with plasma lipoprotein levels: results from the Framingham Offspring Study. Arterioscler Thromb Vasc Biol. 1994; 14: 1105–1113.
Menzel HJ, Kladetzky RG, Assmann G. Apolipoprotein E polymorphism and coronary heart disease. Arteriosclerosis. 1983; 3: 310–315.
Enholm C, Lukka M, Kuusi C, Nikkila E, Utermann G. Apolipoprotein E polymorphism in the Finnish population: gene frequencies and relation to lipoprotein concentrations. J Lipid Res. 1986; 27: 227–235.
Ordovas JM, Litwack-Klein L, Wilson PW, Shaefer MM, Schaefer EJ. Apolipoprotein E isoform phenotyping methodology and population frequency with identification of apoE1 and apoE5 isoforms. J Lipid Res. 1987; 28: 371–380.
Srivastava RAK, Srivastave N, Averna M, Lin RC, Korach S, Lubahn DB, Schonfeld G. Estrogen up-regulates apolipoprotein E (ApoE) gene expression by increasing apoE mRNA in the translating pool via the estrogen receptor α-mediated pathway. J Biol Chem. 1997; 272: 33360–33366.
Rosenfeld ME, Butler S, Ord VA, Lipton BA, Dyer CA Curtiss LK, Palinski W, Witzum JL. Abundant expression of apolipoprotein E by macrophages in human and rabbit atherosclerotic lesions. Arterioscler Thromb Vasc Biol. 1993; 13: 1382–1389.