Association of Inflammatory Gene Polymorphisms With Ischemic Stroke in a Chinese Han Population

Nan Zhao; Xin Liu; Yongqin Wang; Xiaoqiu Liu; Jiana Li; Litian Yu; Liyuan Ma; Shuyu Wang; Hongye Zhang; Lisheng Liu; Jingbo Zhao; Xingyu Wang


J Neuroinflammation. 2012;9(162) 

In This Article

Materials and Methods

Study Participants

The Stroke Hypertension INvestigation IN Genetics (SHINING) study was conducted by Beijing Hypertension League Institute from 1997 to 2000. SHINING study comprised of subjects exclusive to Han ethnicity. A total of 3,119 participants (1,559 stroke cases and 1,560 controls) were recruited from 6 geographical regions in northern China. Cases were recruited from the community of those who were discharged from hospitals between 1997 to 2000; any stroke patients who suffered a stroke within the past 5 years were eligible to participate the study. All patients had to have medical records with diagnosis from brain computed tomography (CT)/MRI. Control subjects were selected according to the case–control study criteria during the same period (control subjects matched to cases by sex, age within 3 years, geographic location, and blood pressure category (<140/90, ≥140/90 and ≤180/105, >180/105 mmHg)). Details of the protocol have been described elsewhere.[9] Cases with a prior history of myocardial infarction as well as controls with a previous history of myocardial infarction or stroke were excluded.

In this study, we focused only on ischemic stroke (1,124 ischemic stroke cases and 1,163 controls) since it is the most prevalent form of stroke that accounts for more than three-quarters of all cases. Data collected included age, sex, body mass index (BMI), systolic blood pressure (SBP), diastolic blood pressure (DBP) and hypertension. Blood pressures were either measured (one to three measurements) or extracted from medical records. Hypertension was defined as systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg or using of antihypertensive medication.

All participants had signed written informed consent and the study was approved by the ethics committee of Beijing Hypertension League Institute.

Selection of SNPs and Genotyping

DNA was extracted from the whole blood with salting out procedure. We used a polymerase chain reaction (PCR)-based inflammatory marker panel (Roche Molecular Biochemicals, Basel, Switzerland) to genotype DNA samples. Basic information on the selected SNPs is shown in Additional file 1: Table S1. These SNPs are not in linkage disequilibrium. The linkage disequilibrium calculation was performed using Haploview 4.1 software. The SNPs on the strip system represent a selection of genetic variants, essentially all of which had been associated with inflammatory diseases before the development of the strips. The detailed genotyping procedure was described previously.[9] In brief, multilocus PCR was carried out in a single tube containing 51 pairs of biotinylated primers. Each amplified PCR product was hybridized with sequence-specific oligoprobes immobilized on a nylon membrane strip; biotin-based color was developed and captured with a scanner and proprietary software developed by Roche Molecular Systems. Genotyping calls were made by two researchers to yield unanimous results. The call rates of SNPs were more than 98%.

Statistical Analysis

Continuous variables were expressed as mean ± SD and differences between case and control groups were analyzed by Student's t test. Categorical variables were presented as percentage. The χ 2 or exact test was used to identify significant departure from Hardy-Weinberg equilibrium. Polymorphisms with a minor allele frequency (MAF) <5% would be excluded from the analyses.

Logistic regression models were used to estimate odds ratios (ORs) and their 95 percent confidence intervals (95% CIs) by using an additive genetic model. An additive model was built that assumed that the risk for carriers of the heterozygous genotype for developing the outcome was half way between carriers of the homozygous genotypes. The advantage was that the strength of genotype-phenotype association was expressed in a single parameter (best estimate) and statistical tests had only 1 degree of freedom.[10] Also this model would have sufficient power to capture dominant modes of transmission given our sample size. In a first step, age-adjusted and sex-adjusted models were constructed. Thereafter, we fitted multivariable models adjusting for age, sex, BMI and hypertension status. Given the similarity of the results, only multivariable adjusted models are presented in this manuscript. To adjust for multiple hypothesis testing, the false discovery rate (FDR) was applied.[11] The FDR significance threshold was defined with a value of 0.20,[12] meaning that one should expect at most 20% of declared discoveries to be false. A multivariate analysis was performed to find interactions between polymorphisms and hypertension status, adjusting for age, sex, BMI. All analyses were performed using SAS V.9.1.3 (SAS, Cary, NC, USA).

Power calculations were generated using a log-additive model of risk, 5% type 1 error rate. For a minor allele frequency of 0.05, the power in the discovery phase of total group was 86.46%; the power in the hypertension group was 65.50%; and in non-hypertension the power was 57.34%. Power calculations were performed using Quanto software (