Patients and Controls. Group A consisted of 306 male patients aged less than 55 years (45.27 ± 0.6 years) with a diagnosis of myocardial infarction (MI), attending the Republic Cardiac Center and Republic Clinical Hospital in Ufa, Bashkortostan, Russia. This diagnosis was based on clinical, enzymatic, electrocardiographic and echocardiographic criteria. All of the following were necessary to confirm the diagnosis: typical chest pain for longer than 30 min., an increase in plasma activity of creatine kinase of more than twice the baseline level, characteristic electrocardiographic changes with Q waves and signs of left ventricular wall akinesia or hypokinesia. Group B consisted of 149 deceased males, aged 53.34 ± 0.8 years, autopsied at the Bureau of Judicial Medical Examination of Bashkortostan, Ufa, Bashkortostan, Russia, because of sudden unexpected death due to coronary artery disease (CAD) without other cardiac signs or symptoms. The control group consisted of 245 unrelated healthy male blood donors, aged 39.2 ± 0.8 years.
Genotyping. Genomic DNA was extracted from whole venous blood by phenol-chloroform extraction [13]. The TNF-α genotypes were determined by polymerase chain reaction (PCR) and NcoI restriction mapping as described in [9] The primers: 5'-AGG CAA TAG GTT TTG AGG GCC AT-3' and 5'-TCC TCC CTG CTC CGA TTC CG-3' were used to amplify a DNA fragment of 107 bp that contains the variable –308 nucleotide in the promoter area of the TNF-α gene. Allele TNFA1 was digested by the restrictase and gave fragments of 87 and 20 bp. The TNFA2 allele was not digested by the restrictase and produced only a 107 bp fragment. The alleles were visualized after electrophoresis in polyacrylamide gel after ethidium bromide staining.
Statistical Analysis. The genotype frequencies were checked for deviation from the Hardy-Weinberg equilibrium by the c2 test, using the Rows x Columns program [14]. Confidence interval (CI) of the genotype frequencies were obtained on the basis of Fisher’s distribution. Pair-wise comparison of genotypes and alleles was done with Fisher’s two-side exact test [15]. The strength of association between the alleles and genotypes was estimated by the odds ratio (OR) [16]. Differences were considered significant at a value of p <0.05.