Baseline clinical characteristics of the CI and control groups
Baseline clinical characteristics of the patients with CI and control groups are summarized in Table 1. A total of 517 Chinese patients with CI, consisting of 344 males and 173 females aged 26-91 years (mean=61.52 years, SD=11.24), and 517 controls participants, comprising 332 males and 185 females aged 23-93 years (mean=61.26 years, SD=14.29), were included in the present study. Most of the study participants originate from northwest China. Age and sex distribution were not significantly different between the two groups (P=0.742 and 0.472, respectively). The proportions of drinking (P=0.002), hypertension (P=0.003) and diabetes (P<0.001) were significantly higher in the CI group than in the control group. TG level was significantly higher in the CI group compared to the control group (P=0.011), no significant differences were found in TC, high-density lipoprotein (HDL), LDL-c, creatinine (Cr), ApoA-I, ApoB, ApoB/A-I, ApoE, lipoprotein, direct bilirubin (DB) and indirect bilirubin (IB) levels between the two groups (all P > 0.05).
Allele And Genotype Frequencies Of Apoe And Ci Risk
The distribution of the ApoE genotype and allele frequencies in CI cases and control subjects is summarized in Table 2. The genotype distribution of this polymorphism in patients with CAD and control participants was in concordance with Hardy-Weinberg equilibrium (P = 0.45 and 0.50, respectively). Genotype ε3/ε3 (63.8%) was the most common type in CI groups, followed by ε3/ε4 (22.8%), ε2/ε3 (9.9%), ε4/ε4 (1.5%), ε2/ε4 (1.2%), and ε2/ε2 (0.8%), whereas those in control participants were ε3/ε3 (71.0%), followed by ε2/ε3 (13.3%), ε3/ε4 (12.6%), ε2/ε4 (2.1%), ε4/ε4 (0.6%) and ε2/ε2 (0.4%). The allele frequency of ε2, ε3, and ε4 was 6.3%, 80.2% and 13.5% respectively in patients with CI; 8.1%, 83.9% and 7.9% respectively in control participants. The distribution of ApoE genotypes and alleles in the two groups was significantly different (χ2=24.424 and P<0.001, χ2=18.472 and P<0.001, respectively).
Table 2
The distributions of genotypes and alleles of the ApoE gene in the CAD patients and controls
Genotype, n (%)
|
CI group(n=617)
|
Control group(n=308)
|
OR (95% CI)
|
Pa
|
χ2
|
P
|
E2/E2
|
4(0.8%)
|
2(0.4%)
|
2.010(0.366,11.010)
|
0.687
|
24.424
|
0.000
|
E2/E3
|
51(9.9%)
|
69(13.3%)
|
0.711(0.484,1.044)
|
0.099
|
E2/E4
|
6(1.2%)
|
11(2.1%)
|
0.540(0.198,1.472)
|
0.328
|
E3/E3
|
330(63.8%)
|
367(71.0%)
|
0.721(0.555,0.937)
|
0.017
|
E3/E4
|
118(22.8%)
|
65(12.6%)
|
2.057(1.477,2.864)
|
0.000
|
E4/E4
|
8(1.5%)
|
3(0.6%)
|
2.693(0.710,10.208)
|
0.224
|
HWE
|
χ2 = 3.70, P = 0.45
|
χ2 = 3.33, P =0.50
|
|
|
|
|
Alleles, n (%)
|
|
|
|
|
|
|
E2
|
65(6.3%)
|
84(8.1%)
|
0.759(0.542,1.062)
|
0.126
|
18.472
|
0.000
|
E3
|
829(80.2%)
|
868(83.9%)
|
0.773(0.617,0.969)
|
0.029
|
E4
|
140(13.5%)
|
82(7.9%)
|
1.818(1.364,2.424)
|
0.000
|
CI: cerebral infarction, HWE: Hardy-Weinberg equilibrium. |
ap and OR (95% CI) values were calculated by logistic regression adjusted for age, gender, and traditional cardiovascular risk factors. |
bp values were calculated from two-sided chi-square tests or Fisher’s exact tests. |
The frequencies of ε3/ε4 genotype (OR =2.057, 95% CI = 1.477–2.864, P<0.001) and ε4 allele (OR =1.818, 95% CI = 1.364–2.424, P<0.001) were significantly higher in CI patients than in control participants. Further, patients with CAD had a significantly lower ε3/ε3 (OR =0.721, 95% CI = 0.555–0.937, P=0.017) genotype and ε3 allele (OR =0.773, 95% CI = 0.617–0.969, P=0.029) frequencies than did the control participants. (P < 0.05)
To explore the relationship between ApoE genotype and CI, we conducted further analysis stratified by age(dichotomized into ≤60 years and >60 years) and sex. The results showed that ε3/ε4 frequency was significantly higher in patients with CI compared to the control participants (OR =3.067, 95% CI = 1.675–5.614, P<0.001 in age ≤60 years; OR =1.735, 95% CI = 1.156–2.604, P=0.008 in age >60 years and OR =2.206, 95% CI = 1.474–3.301, P<0.001 in males), but not in females (OR =1.746, 95% CI = 0.973–3.134, P=0.078). Additionally, the variance in allele ε4 between patients with CI and controls was also statistically significant (OR =2.072, 95% CI = 1.281–3.353, P=0.003 in age ≤60 years; OR =1.704, 95% CI = 1.189–2.444, P=0.003 in age>60 years; OR =1.709, 95% CI = 1.201–2.432, P=0.001 in males and OR =2.046, 95% CI = 1.246–3.361, P=0.004 in females) (Table 3).
Relationships between serum lipid profile and ApoE alleles.
Table 4 describe the association between serum lipid profiles and allelic carrier status (ε2, ε3 and ε4 groups). Participants with ε2/ε4 genotype (n= 17) were excluded because play opposing roles in lipid metabolism and the incidence of CI. In the patients with CAD, ε4 carriers had significantly higher LDL-C, ApoB and ApoB/ApoA-I and lower levels of ApoA-I and ApoE levels than ε2 carriers. LDL-C, ApoB, ApoB/ApoA-I and ApoE levels of the control participants showed similar trends to those in CI groups. Additionally, control participants with ε4 carriers had significantly higher levels of lipoprotein and TC levels than ε2 carriers.
Table 4
Relationships between serum lipid profile and ApoE allele in CI patients and control participants
CI patient
|
Control participants
|
Factors
|
ε2(ε2ε2 + ε2ε3)
|
ε3(ε3)
|
ε4(ε3ε4 + ε4ε4)
|
P a
|
P b
|
P c
|
P d
|
ε2(ε2ε2 + ε2ε3)
|
ε3(ε3)
|
ε4(ε3ε4 + ε4ε4)
|
P a
|
P b
|
P c
|
P d
|
TC
|
3.59±0.87
|
3.86±0.94
|
3.78±0.90
|
0.112
|
0.046
|
0.423
|
0.183
|
3.55±0.87
|
3.82±0.84
|
3.85±0.93
|
0.043
|
0.014
|
0.822
|
0.042
|
HDL
|
1.07±0.26
|
1.02±0.24
|
1.00±0.27
|
0.220
|
0.208
|
0.329
|
0.107
|
1.01±0.21
|
1.04±0.27
|
1.04±0.27
|
0.620
|
0.326
|
0.923
|
0.466
|
LDL-C
|
1.89±0.54
|
2.27±0.82
|
2.18±0.79
|
0.004
|
0.001
|
0.266
|
0.014
|
1.98±0.77
|
2.23±0.71
|
2.29±0.77
|
0.023
|
0.009
|
0.589
|
0.043
|
TG
|
1.55±1.23
|
1.45±0.80
|
1.61±1.16
|
0.245
|
0.435
|
0.090
|
0.740
|
1.55±0.84
|
1.36±0.76
|
1.23±0.60
|
0.039
|
0.057
|
0.184
|
0.011
|
Cr
|
64.91±20.61
|
67.69±36.85
|
68.36±34.85
|
0.854
|
0.622
|
0.870
|
0.537
|
65.53±21.07
|
70.03±60.95
|
69.91±26.78
|
0.820
|
0.553
|
0.987
|
0.302
|
Apolipoprotein A
|
1.26±0.23
|
1.16±0.20
|
1.12±0.21
|
0.000
|
0.002
|
0.061
|
0.000
|
1.17±0.18
|
1.16±0.21
|
1.13±0.22
|
0.514
|
0.710
|
0.139
|
0.263
|
Apolipoprotein B
|
0.64±0.14
|
0.77±0.20
|
0.77±0.21
|
0.000
|
0.000
|
0.944
|
0.000
|
0.66±0.21
|
0.75±0.20
|
0.76±0.21
|
0.001
|
0.000
|
0.535
|
0.003
|
ApoB/ApoA
|
0.54±0.14
|
0.68±0.21
|
0.70±0.20
|
0.000
|
0.000
|
0.274
|
0.000
|
0.56±0.20
|
0.67±0.31
|
0.68±0.21
|
0.012
|
0.005
|
0.807
|
0.001
|
Apolipoprotein E
|
49.03±21.64
|
34.61±10.89
|
32.04±14.05
|
0.000
|
0.000
|
0.039
|
0.000
|
50.24±17.97
|
34.99±11.75
|
32.11±12.09
|
0.000
|
0.000
|
0.066
|
0.000
|
Lipoprotein
|
215.81±213.97
|
210.85±207.31
|
212.14±201.12
|
0.986
|
0.870
|
0.952
|
0.912
|
133.86±109.56
|
203.59±216.48
|
213.16±182.36
|
0.021
|
0.008
|
0.732
|
0.002
|
DB
|
4.58±2.26
|
4.45±2.25
|
4.58±2.74
|
0.863
|
0.729
|
0.626
|
0.990
|
4.73±2.50
|
4.65±5.19
|
5.14±4.14
|
0.737
|
0.904
|
0.460
|
0.472
|
IB
|
9.08±5.26
|
9.14±4.58
|
9.13±5.12
|
0.998
|
0.943
|
0.991
|
0.959
|
9.14±4.41
|
9.06±4.79
|
9.78±5.04
|
0.526
|
0.890
|
0.263
|
0.434
|
CI: cerebral infarction, TC: total cholesterol, HDL: high-density lipoproteins, LDL-C: low-density lipoprotein cholesterol, TG: triglyceride, Cr: creatinine, DB: direct bilirubin, IB: indirect bilirubin |
ap value shows the differences compared between groups (ε2, ε3, ε4) |
bp values obtained when comparing ε2 subjects with ε3 subjects. |
cp values obtained when comparing ε4 subjects with ε3 subjects. |
dp values obtained when comparing ε2 subjects with ε4 subjects. |
Logistic Regression Analysis Of Ci Risk Factors
We performed a multivariate logistic regression analysis to identify which variables with statistical significance from the univariate analysis could act as independent predictors of CI. Univariate logistic analysis showed that drinking (OR =1.701, 95% CI = 1.167–2.478, P=0.006), hypertension (OR =1.885, 95% CI = 1.417–2.508, P<0.001), ApoE levels (OR =0.983, 95% CI = 0.969–0.997, P=0.017), lipoprotein (OR =1.001, 95% CI = 1.000–1.001, P=0.046), TG (OR =1.363, 95% CI = 1.035–1.797, P=0.028) and ε4 allele (OR =1.954, 95% CI = 1.359–2.810; P<0.001) were significant independent risk factors for CI (Table 5). Furthermore, to exclude the confounding effect of age (≤60 years and >60 years) and sex, multivariate logistic regression analysis stratified according to age and sex was performed. In both age groups, ε4 carriers were associated with increased risk of CI (age ≤60 years: OR = 2.970, 95% CI = 1.553–5.678, P = 0.001; age >60 years: OR = 1.715, 95% CI = 1.082–2.719, P = 0.022). Also, ε4 carriers was significantly associated with a higher risk of CI in males (OR =2.182, 95% CI = 1.398–3.407, P=0.022), but not in females (OR =1.500, 95% CI = 0.779–2.887, P=0.225) (Table 6).
Table 5
Logistic regression analysis of the risk of CI in Northwest of China population.
Variables
|
P- value
|
OR (95% CI)
|
Drinking
|
0.006
|
1.701(1.167-2.478)
|
Hypertension
|
0.000
|
1.885(1.417-2.508)
|
Lipoprotein
|
0.046
|
1.001(1.000-1.001)
|
TG
|
0.028
|
1.363(1.035-1.797)
|
ApoE
|
0.017
|
0.983(0.969-0.997)
|
ε4
|
0.000
|
1.954(1.359-2.810)
|
CI: cerebral infarction, TG: triglyceride |
Table 6
Multiple logistic regression analysis for CI patients and control subjects
ε4
|
|
OR (95% CI)
|
P-value
|
Age≤60
|
2.970(1.553-5.678)
|
0.001
|
Age༞60
|
1.715(1.082-2.719)
|
0.022
|
males
|
2.182(1.398-3.407)
|
0.001
|
females
|
1.500(0.779-2.887)
|
0.225
|