A case-reference study on plasma fibrinogen concentrations and coronary atherosclerosis in Japanese.

To examine contribution of plasma fibrinogen concentrations to coronary atherosclerosis in Japanese, a cross-sectional study was conducted from 1991 to 1993 for 169 myocardial infarction cases and 1 : 1 matched references. All cases had one to three coronary vessels with 50 percent or more stenosis. References were chosen out of Osaka residents who had no cardiac event by matching sex and age ((+)-3 years) with cases. Mean (SD) values of plasma fibrinogen concentration were 322 (81) mg/dl in men and 377 (77) mg/dl in women for cases, and were 282 (56) mg/dl in men and 277 (48) mg/dl in women for references; mean plasma fibrinogen was significantly higher in cases than in references for both men and women. Furthermore, mean plasma fibrinogen was progressively higher as the number of stenotic vessels increased for men, and similar trend was seen for women. The case-reference difference remained significant even after adjusting smoking status, hypertension, serum total cholesterol, serum HDL cholesterol. Although plasma fibrinogen concentration was reported lower in Japanese than in American Caucasians, plasma fibrinogen is suggested to be a risk factor for myocardial infarction among Japanese as so among Caucasians.

Hypertension, smoking and hypercholesterolemia are three major risk factors for incidence and mortality from coronary heart disease1,2). Recently, in addition to these, a lower serum HDL cholesterol concentration 3-6) and a higher plasma fibrinogen concentration 7-12) are recognized independent risk factors. According to our follow-up studies in urban Japanese men, hypertension, smoking, hypercholesterolemia13) and a lower serum HDL cholesterol concentration14) have been confirmed coronary risk factors. Regarding plasma fibrinogen, we reported positive associations of plasma fibrinogen concentrations with age, blood pressure, cigarette smoking and serum total cholesterol concentrations, and inverse association with serum HDL cholesterol concentrations in urban Japanese men15). These correlations between plasma fibrinogen concentrations and known coronary risk factors suggest that higher plasma fibrinogen concentrations increase the risk of coronary heart disease in Japanese as so in Caucasians. As the first step to confirm this hypothesis, we conducted a case-reference study to examine the relation between plasma fibrinogen concentrations and coronary atherosclerosis. We used the analysis of covariance for mean values, and chi-square test for proportions to examine differences between cases and references, and differences according to the number of vessels with 75% or more stenosis. Conditional multivariate logistic regression analysis was conducted to examine the relation of plasma fibrinogen concentrations and other related variables with the prevalence of myocardial infarction. All probability values for statistical tests were two-tailed.

RESULTS
Sex-specific cardiovascular risk factors are shown in Table  1. Mean plasma fibrinogen concentration was higher in cases than in references for both sexes. Mean serum total cholesterol concentration was higher in cases than in references for men, but did not differ between cases and references for women. Mean serum HDL cholesterol concentration was lower in cases than in references for both sexes. There was no significant difference in either systolic or diastolic blood pressure between cases and references.
However, the proportion of hypertensive persons was significantly higher in cases than in references for both sexes. Mean number of cigarettes per day and the proportion of smokers were higher in cases than in references for both sexes. There was no significant difference between cases and references in mean alcohol intake or the proportion of drinkers for either sexes. Sex specific findings of the cases according to the number of stenotic vessels, compared to references, were shown in Tables 2 and 3. Mean plasma fibrinogen concentration was progressively higher as the number of stenotic vessels incresed for men, and a similar trend was seen for women. Mean serum total cholesterol concentration was also progressively higher with the number of stenotic vessels for men, but did not vary significantly for women. Mean serum HDL cholesterol concentration was lower in cases than in references, but did not  Stenotic coronary vessels were having 75% or more stenosis, three cases having less than 75% stenosis were excluded from the analysis. Differences among references and three subgroupes of cases *:p<0.05 **:p<0.01 Table 3. Mean (SD) values and prevalence of coronary risk factors according to the number of stenotic coronary vessels for women Stenotic coronary vessels were having 75% or more stenosis, three cases having less than 75% stenosis were excluded from the analysis.
Differences among references and three subgroupes of cases *:p<0.05 **:p<0.01 ly related with myocardial infarction. A 64mg/dl higher plasma fibrinogen concentration was associated with two times higher risk of myocardial infarction.

DISCUSSION
In this study we found plasma fibrinogen was higher in cases than in references for both men and women. This relation remained significant even after adjusting smoking status, hypertension, serum total cholesterol and serum HDL cholesterol. Previous cross-sectional studies in Western countries recognized the relation between plasma fibrinogen and coronary heart disease or coronary atherosclerosis22-26). In Japan, one cross-sectional case study reported the relation between plasma fibrinogen concentrations and the number of stenotic vessels : the relation was statistically significant in men but not in women27). In the present study, however, we found the statistically significant relation both in men and women.
There are two limitations since this study is based on a cross-sectional design. The first limitation is from case ascertainment. All cases for this study were survivers who discharged the hospital, fatal or severe nonfatal cases were not examined. On the other hand, the cases had at least one stenotic vessel and we probably missed less severe cases. Thus, it is uncertain whether we selected more or less severe cases which may influence the risk factor analyses. The second limitation is on the timing that we examined cases : one month after the discharge. We possibly underestimated the contribution of risk factors which may change rapidly as patients changed their lifestyles after the onset of myocardial infarction. For instance, serum total cholesterol may decline and serum HDL cholesterol may increase after patients reduced fat intake, reduced body weight or quitted cigarette smoking. Blood pressure was more likely to decrease when patients started antihypertensive medication and changed their lifestyles. Increased plasma fibrinogen concentration among smokers, however, would not decline so rapidly even after the cessation of smoking. Several cross-sectional studies suggested that it takes two or more years after the smoking cessation to reach plasma fibrinogen levels in never-smokers 28-31). Thus, the contribution of plasma fibrinogen concentration to the risk of myocardial infarction was unlikely to be underestimated.
Prospective studies in Western countries7-12) have shown that plasma fibrinogen is a risk factor of myocardial infarction. Our previous cross-sectional study 15) showed that plasma fibrinogen concentrations were positively associated with age, smoking, serum total cholesterol concentration, and inversely associated with serum HDL cholesterol concentration which was consistent with the results of studies in Western countries [9][10][11]29,30,[32][33][34][35][36][37][38][39][40][41][42] Although plasma fibrinogen concentrations were lower in Japanese than in American Caucasians according to our previous cross-cultured study43), the present study suggested that plasma fibrinogen is a risk factor of myocardial infarction among Japanese as so among Caucasians. Our prospec-tive study is underway to confirm this hypothesis.