Coagulation Factor VII and the Risk of Coronary Heart Disease in Healthy Men
Abstract Numerous investigations have demonstrated the role of thrombus formation in the pathogenesis of coronary heart disease (CHD). A tendency to thrombosis may also be indicated by elevated levels of coagulation factor VII clotting activity (FVIIc). Significant associations of FVIIc with increased coronary risk, however, have been found only in the Northwick Park Heart Study. Here we present the results of the 8-year follow-up of FVIIc measurements in 2780 healthy men of the Prospective Cardiovascular Münster study. In the study population (age at entry, 49.3±6.1 years, mean±SD), 130 CHD events occurred during follow-up. FVIIc was significantly higher in subjects with coronary events than in those without (112.4±20.1% vs 108.7±21.4%, P=.023). Compared with individuals without coronary events, FVIIc was not significantly higher in men with nonfatal events (111.7±20.4%; P=.196, n=93), but there was a tendency toward higher FVIIc activity in subjects with fatal events (114.6±19.5%; P=.076, n=37). In the multiple logistic regression analysis, we did not find FVIIc to be an independent risk factor for CHD, and the significance of FVIIc disappeared after total cholesterol, LDL-cholesterol, and triglycerides were taken into account. The increase in the number of CHD events through higher levels of FVIIc was more pronounced in the presence of additional cardiovascular risk factors: smoking; myocardial infarction events in family; angina pectoris; high levels of fibrinogen, total cholesterol, LDL cholesterol, and triglycerides; and a low level of HDL cholesterol. We conclude that FVIIc is a risk factor for CHD, especially in the presence of additional risk factors, and must be taken into account when assessing cardiovascular risk in men.
- Received August 2, 1996.
- Accepted October 21, 1996.
Diseases of the cardiovascular system are the major cause of morbidity and mortality in Western countries. The development of atherosclerosis, MI, and sudden cardiac death is at least partly due to increased activity of the coagulation system. Several investigations have demonstrated the role of thrombus formation in the pathogenesis of atherosclerosis and acute cardiovascular events.1 2 3 4 The NPHS has indicated that FVIIc is independently associated with the risk of CHD.5 In addition, in several cross-sectional studies, elevated plasma levels of FVIIc have been shown to be associated with CHD.6 7 8 Nevertheless, not all studies have confirmed FVIIc as an independent risk factor. Hence, the significance of FVIIc in CHD remains unclear.9 10
In 1994 we published the results of the 6-year follow-up of the PROCAM Study, which showed that there was a tendency for plasma levels of FVIIc to be higher in subjects who had experienced a CHD event.11 Here we present the results of the 8-year follow-up of FVIIc measurements in 2780 healthy men.
Study Design and Subjects
The PROCAM Study started in 1979. Apparently healthy employees of Westphalian companies were examined specifically for cardiovascular risk factors and then kept under observation to record mortality, subsequent MI, and stroke. The initial baseline examination included each participant’s medical history, physical examination, ECG at rest, and a comprehensive laboratory blood analysis. In 1981 determination of plasma fibrinogen and FVIIc levels was included. A total of 10 581 subjects (7540 men and 3041 women) were recruited up to 1985. From 1981 onward, significant numbers of MIs or CHD deaths occurred only in men aged 40 and older. Statistical analysis was therefore confined to the 2780 male participants between 40 and 65 years of age with no prior history of MI or stroke who had completed the 8-year follow-up (Table 1⇓).
Diagnostic criteria and the definition of end points have been published in detail.12 Two end points were considered: definite nonfatal MI and definite CHD death, including sudden cardiac death and fatal MI. Definite nonfatal MI was diagnosed if one or more of the following conditions were fulfilled: diagnostic ECG at the time of the event; ischemic cardiac pain plus diagnostic enzyme levels; ischemic cardiac pain plus equivocal enzyme levels and equivocal ECG; or an ECG that was diagnostic for MI while a previous one was not. Sudden cardiac death was diagnosed if an apparently well subject was observed to have died within 1 hour of onset of symptoms or within 1 hour of having last been seen without them and the cause of death could not be reasonably attributed (on the basis of complete clinical information) to violence, trauma, or some potentially lethal disease other than CHD. Fatal MI was diagnosed when a death certificate or hospital record described the cause of death accordingly or when an autopsy finding of acute MI was available. The members of the Critical Event Committee (see “Appendix”) verified the diagnoses and causes of death in all event cases.
Blood was drawn by venipuncture in the morning after a 12-hour fast. Blood samples for coagulation tests were collected into plastic tubes containing 1/10 of the volume with 3.13% trisodium citrate. Within 60 minutes of blood collection, the plasma was separated by centrifugation at room temperature for 15 minutes at 2500g. Serum from blood samples was prepared without additives for clinical chemistry by centrifugation for 10 minutes at 3000g. After aliquots were placed in plastic tubes, plasma and serum were kept in liquid N2 until they were stored in the laboratory at −70°C.
FVIIc and fibrinogen were measured on a KC10 coagulation analyzer (Amelung). FVIIc was determined in a one-step assay using thromboplastin (Thromborel S) and factor VII–deficient plasma (both from Behringwerke). Fibrinogen was determined according to the Clauss method using thrombin (Multifibren), control plasma (both from Behringwerke), and pooled plasma. Measurement of TGs and total cholesterol in serum was performed on a Hitachi 737 autoanalyzer (Boehringer Mannheim). HDL-C concentrations were determined after precipitation with phosphotungstic acid/MgCl2 (Boehringer Mannheim), and LDL-C was calculated using the Friedewald formula.
Continuous variables were compared by the t test (TGs after logarithmic transformation), and discrete parameters were compared by the χ2 test. In all analyses except where age was included as an independent variable in the multiple logistic regression, results were adjusted for age by using regressions derived from the entry data. For additional description of the relation between a specific variable and the probability of suffering a CHD event, the number of CHD events in tertiles of the variable is given. In addition, a multiple logistic regression analysis was performed; the variables included age, systolic blood pressure, diabetes mellitus, smoking, MI in the family, angina pectoris, total cholesterol, LDL-C, HDL-C, and TGs. ORs were calculated for the lower and upper tertiles of variables. The statistical analysis was done with the software packages spssx and sas.
Incidence and Mortality
In the 2780 subjects studied (age at entry, 49.3±6.1 years; range, 40 to 65), a total of 130 CHD events (15 sudden cardiac deaths, 22 fatal MIs, and 93 nonfatal MIs) occurred during the observation period of 96 months, which corresponds to an annual incidence of 0.62%. Sixty-nine patients died from causes other than CHD (8 from other diseases of the circulatory system, 33 from malignant neoplasms, 16 from accidents and violence, and 12 from other diseases). In addition, 20 nonfatal strokes were observed (Table 2⇓).
Comparison of Subjects With and Without CHD Events
After adjustment for age, mean values for fibrinogen, FVIIc, total cholesterol, LDL-C, the LDL-C to HDL-C ratio, TGs, and systolic and diastolic blood pressure were revealed to be significantly higher in the event group, whereas HDL-C was higher in subjects without CHD events. There was no significant difference in BMI between the two groups (Table 3⇓).
Incidences by tertiles of variables are demonstrated in Table 4⇓. The strongest associations were observed for tertiles of the LDL-C to HDL-C ratio (OR, 6.3) and LDL-C (OR, 4.9). The weakest association was found for BMI (OR, 1.5) and FVIIc (OR, 1.6).
Comparison of Subjects With Fatal and Nonfatal CHD Events and Those Without CHD Events
As shown in Table 5⇓, no significant differences in fibrinogen were found between subjects with nonfatal (mean fibrinogen level, 2.91±0.57 g/L) and fatal (mean fibrinogen level, 2.92±0.64 g/L) CHD events. Compared with the no-event group (mean fibrinogen level, 2.59±0.57 g/L), fibrinogen was significantly higher in patients with nonfatal (P<.001, n=93) as well as fatal (P=.004, n=37) events.
For FVIIc, there was a tendency to higher FVIIc activities in individuals with fatal events after the CHD group was divided into subgroups with nonfatal versus fatal events. The mean FVIIc value was 111.7±20.4% in subjects with nonfatal CHD events and 114.6±19.5% in subjects with fatal events. Compared with subjects without any coronary events, FVIIc (mean, 108.7±21.4%) was not significantly higher in patients with nonfatal events (mean, 111.7±20.4%; P=.196, n=37), but a borderline significant difference appeared when the no-event group was compared with the fatal-event group (mean, 114.6±19.5%; P=.076, n=93).
Results of Multiple Logistic Regression Analysis
To consider the independence of FVIIc as a predictor of CHD, a multiple logistic regression analysis was performed. As shown in Table 6⇓, FVIIc remained a significant predictor of CHD events after adjustment for age, systolic blood pressure, smoking, diabetes mellitus, angina pectoris, MI events in the family, fibrinogen, and HDL-C. After total cholesterol, LDL-C, and TGs were taken into account, however, this significance was lost.
Incidence of CHD Events for Combined Tertiles of FVIIc and Nonlaboratory Variables
The combination of nonlaboratory parameters (systolic blood pressure, smoking habit, diabetes mellitus, angina pectoris, and MI in the family) and tertiles of FVIIc and the resulting incidence are summarized in Table 7⇓. On the one hand, for subjects with MI in the family or angina pectoris and for smokers, there was a higher OR with increasing levels of FVIIc than for subjects without the additional risk factor (smoker-to-nonsmoker OR, 2.1/1.3; with-to-without MI in the family OR, 3.6/1.4; angina pectoris-to-no angina pectoris OR, 2.4/1.6). On the other hand, the coronary risk of smokers and subjects with MI in the family was more pronounced in the upper than the lower tertile of FVIIc (ORs of the upper-to-lower tertiles of FVIIc: for smokers, 2.5/1.6; for those with MI in the family, 1.8/0.7; and for those with angina pectoris, 6.8/4.5).
Incidence of CHD Events for Combined Tertiles of FVIIc and Other Laboratory Variables
Incidences for combinations of tertiles of FVIIc and fibrinogen, total cholesterol, LDL-C, HDL-C, and TGs are summarized in Table 8⇓. For all parameters the OR was higher within the upper than the lower tertile of the corresponding risk factor (although vice versa for HDL-C). The ORs of FVIIc varied between 1.1 (within the lower tertiles of total cholesterol and TGs and the lower and middle tertiles of LDL-C) and 1.9 (within the upper tertile of fibrinogen and the lower tertile of HDL-C). On the other hand, the importance of fibrinogen, total cholesterol, LDL-C, HDL-C, and TGs as risk factors for CHD was greater in the presence of high levels of FVIIc (ORs within the upper to lower tertiles of FVIIc: for fibrinogen, 2.8/2.0; for total cholesterol, 3.2/2.4; for LDL-C, 6.3/4.5; for HDL-C, 2.9/2.0; and for TGs, 2.0/1.4).
As displayed in the Figure⇓, the highest incidence values (111.8 per 1000 men) were due to the combination of high FVIIc and LDL-C. When compared with the lower tertiles of both parameters (incidence, 16.4 per 1000 men), this value is a nearly 7-fold increase in risk. The lowest number of events within the upper tertiles and the lowest increase in risk compared with the low tertiles were found for FVIIc and TGs (lower tertiles, 35.4 per 1000 men; upper tertiles, 79.0 per 1000 men; 2.2-fold increase).
The results delineated in this article confirm and extend the results of the 6-year follow-up of the PROCAM Study. The annual incidence was nearly unchanged (0.65% versus 0.62%) and low when compared with the results of the NPHS (0.8%, 5-year follow-up).5 11 Our recent results confirm fibrinogen as an independent risk factor for CHD.13 Owing to a longer observation period and a greater CHD incidence, FVIIc now also appears to be significantly higher in subjects suffering coronary events than in those without such events (mean FVIIc, 112.4±20.1% versus 108.7±21.4%; P<.05).
However, as shown in Table 4⇑, the incidence of CHD was 1.6-fold higher in the upper tertile of FVIIc than in the lower tertile. Compared with the other risk factors, eg, LDL-C with an OR of 4.9, 1.6 represents only a slight increase in incidence. In the NPHS, FVIIc was found to be much more strongly associated with coronary risk (117.4% versus 107.0%; P<.001, 5-year follow-up).5
As shown by the results of a laboratory comparison, the FVIIc assay used in the NPHS is more sensitive to the activated form of coagulation factor VII (ie, FVIIa) than is the PROCAM assay. Plasma samples enriched with FVIIa expressed a proportionately greater FVIIc in the NPHS assay using mixed human and bovine thromboplastin than in the PROCAM pure human thromboplastin system. It was also true that the converse held in samples that were relatively deficient in FVIIa. Therefore, the clear-cut association of FVIIc with CHD in the NPHS may be due to the ability of their assay to detect increased circulating levels of the two-chain form of FVII.14 To our knowledge, there are no differences between the NPHS and PROCAM sample collecting protocols that may have contributed to the discrepant results concerning FVIIc.
Table 5⇑ displays the results for MI survivors and those who died of MI or sudden cardiac death. In contrast to the entire group, the significance of FVIIc was lost within the subgroups because of fewer individuals. Nevertheless, differences in FVIIc were still found. FVIIc showed a borderline significance when individuals with fatal events were compared with subjects without events (114.6% versus 108.7%, P=.076). In subjects with nonfatal events, FVIIc was not significantly different from both of the other groups (111.7%, P=.169 and .451, respectively). Contrastingly, the difference in FVIIc between subjects suffering fatal events and those suffering nonfatal events was significant in the NPHS (P=.008).5 However, both sets of results support the assumption of an activated coagulation system, which increases the risk of fatal events due to the enhanced size and stability of the occlusive thrombus.15
After performing multiple logistic regression analysis, we could not confirm FVIIc as an independent risk factor for CHD, as demonstrated in the NPHS.5 As shown in Table 6⇑, no significance was found after accounting for total cholesterol, LDL-C, and TGs. This finding was confirmed by the data displayed in Table 8⇑ and the Figure⇑: within the lower and the middle tertiles of LDL-C, the increase in incidence through higher levels of FVIIc was negligible. These findings were similar but less pronounced for the combination of tertiles of FVIIc and tertiles of total cholesterol and TGs. In contrast, after adjustment for age, fibrinogen, HDL-C, and nonlaboratory parameters (systolic blood pressure, angina pectoris, smoking, diabetes mellitus, and MI events in the family), FVIIc remained significant.
Activation of FVII and the subsequent increase in FVIIc is partly due to an elevation of the hemostatic balance by interactions between lipoproteins and coagulation factors. A correlation between FVIIc levels and TGs as well as total cholesterol has been shown in several studies.16 17 18 19 20 21 Dietary fat intake increased FVIIc, whereas treatment of hypertriglyceridemia with lipid-lowering drugs resulted in a decrease of FVIIc.18 19 On the one hand, higher levels of FVIIc in hyperlipidemic patients may be due to enhanced synthesis of FVII; on the other hand, this may be due to a larger proportion of the activated form of FVIIc.17 21 The correlation between FVIIc and lipoproteins may be explained by the binding of FVII to TG-rich particles.22 23 However, these results support our finding that FVIIc is not independently associated with coronary risk. Higher ORs for FVIIc in the presence of other laboratory and nonlaboratory risk factors (and vice versa) may be explained by additive effects of the two risk factors on atherosclerotic progress. Moreover, direct interaction between the coagulation system and lipids contributes to the elevation of risk for CHD.
Subjects included in this study were apparently healthy. Nevertheless, already-existing minimal atherosclerotic lesions could not be excluded. This factor may be especially relevant for subjects with high levels of laboratory risk factors as well as the nonlaboratory risk factors. The release of tissue factor during rupture of an atheromatous plaque is associated with CHD by means of a dramatically enhanced activation of coagulation factor X by FVIIa in the presence of tissue factor. Therefore, even minimal lesions of atherosclerosis may lead to activation of the coagulation system, which contributes to atherosclerotic progress.15 24 25 26 The assumption that the coagulation system plays a major role in thrombus formation in patients with already-existing CHD is further supported by our findings in patients suffering angina pectoris, who had the highest risk of developing CHD. The OR of FVIIc was higher for subjects suffering from angina pectoris than for those without. Additionally, the OR of angina pectoris was higher within the upper tertile of FVIIc then in the lower tertile. For tertiles of systolic blood pressure and FVIIc, we found the OR of FVIIc to be higher within the lower tertile of systolic blood pressure than the middle and upper tertiles. Furthermore, patients suffering from diabetes mellitus showed a lower OR than did those without. Both findings may be due to the low number of cases and need further investigation.
On the whole, FVIIc is a risk factor for CHD and has to be taken into account when estimating cardiovascular risk in men. The increase in cardiovascular risk through elevated levels of FVIIc depends mainly on the presence of additional risk factors. Consequently, interaction between FVIIc and other risk factors requires further investigation.
Selected Abbreviations and Acronyms
|BMI||=||body mass index|
|CHD||=||coronary heart disease|
|FVII||=||coagulation factor VII|
|FVIIa||=||activated coagulation factor VII|
|FVIIc||=||coagulation factor VII clotting activity|
|NPHS||=||Northwick Park Heart Study|
|PROCAM||=||Prospective Cardiovascular Münster Study|
Members of the Critical Event Committee included K. Kochsiek, MD, Würzburg; B.E. Strauer, MD, Düsseldorf; U. Gleichmann, MD, Bad Oeynhausen; and R. Uebis, MD, Aachen, Germany.
This study was supported by the Bundesministerium für Forschung und Technologie, Ministerium für Wissenschaft und Forschung NRW, Deutsche Forschungsgemeinschaft, Landesversicherungsanstalt Westfalen, and Landesversicherungsanstalt Rheinprovinz. The excellent technical cooperation of J. Bailey, M. Käse, and C. Sandau is gratefully acknowledged.
Davies MJ, Woolf N, Robertson WB. Pathology of acute myocardial infarction with particular reference to occlusive coronary thrombi. Br Heart J. 1976;38:659-664.
Schwartz CJ. Thrombosis in the pathogenesis of sudden cardiac death and myocardial infarction. In: Oates JA, ed. Prostaglandins and the Cardiovascular System. New York, NY: Raven Press; 1982:1-14.
Cortellaro M, Boschetti C, Cofrancesco E, Zanussi C, Catalano M, de Gaetano G, Gabrielli L, Lombardi B, Specchia G, Tavazzi L, Tremoli E, della Volpe A, Polli E, and the PLAT (Progetto Lombardo Atero-Trombosi) Study Group. The PLAT Study: hemostatic function in relation to atherothrombotic ischemic events in vascular disease patients−principal results. Arterioscler Thromb. 1992;12:1063-1070.
Heinrich J, Balleisen L, Schulte H, Assmann G, van de Loo J. Fibrinogen and factor VII in the prediction of coronary risk: results from the PROCAM study in healthy men. Arterioscler Thromb. 1994;14:54-59; Erratum. Arterioscler Thromb. 1994;14:1392.
Miller GJ, Stirling Y, Esnouf MP, Heinrich J, van de Loo J, Kienast J, Wu KK, Morrissey JH, Meade TW, Martin JC, Imeson JD, Cooper JA, Finch A. Factor VII-deficient substrate plasmas depleted of protein C raise the sensitivity of the factor VII bio-assay to activated factor VII: an international study. Thromb Haemost. 1994;71:38-48.
Zitoun D, Bara L, Basdevant A, Samama MM. Levels of factor VIIc associated with decreased tissue factor pathway inhibitor and increased plasminogen activator inhibitor-1 in dyslipidemias. Arterioscler Thromb Vasc Biol. 1996;16:77-81.
de Sousa JC, Soria C, Ayrault-Jarrier M, Pastier D, Bruckert E, Amiral J, Bereziat G, Caen JP. Association between coagulation factors VII and X with triglyceride rich lipoproteins. J Clin Pathol. 1988;41:940-944.
Esnouf MP. Biochemistry of blood coagulation. In: Meade TW, ed. Anticoagulants and Myocardial Infarction. Chichester, UK: J Wiley; 1984:29-62.
Nemerson Y, Zur M, Bach R, Gentry R. The mechanism of action of tissue factor. In: Mann KG, Taylor FB, eds. The Regulation of Coagulation. Amsterdam, Netherlands: Elsevier/North Holland; 1980:193-202.