Relationship Between the Blood Coagulation-Fibrinolysis System and the Subclinical Indicators of Arteriosclerosis in a Healthy Male Population

A cross-sectional observation was performed to assess the relationship between the coagulation-fibrinolysis system and the subclinical indicators of arteriosclerosis in a healthy male population. Subjects were 445 workers (18.9-49.4, Av.36.2yrs) in viscose rayon manufacturing factories in Japan. Coagulation-fibrinolysis parameters determined were D-dimer(DD) , thrombin antithrombinIII complex (TAT), tissue plasminogen activator (TPA),and plasminogen activator inhibitor 1 (PAI1). The following indicators of arteriosclerosis were examined; systolic and diastolic blood pressure (SBP,DBP), stiffness parameter of the carotid artery using ultrasound (3), pulse wave velocity of the aorta (PWV), and a number of lacunar infarctions from brain MRI. After age-stratification(-29,30-39,40+yrs),the subjects were classified into quartiles by coagulation-fibrinolysis parameters. The mean values of SBP and DBP and ,3 and PWV the prevalence of brain infarctions were compared across these quartiles by means of analysis of variance, chi-square test, respectively . Multivariate analysis was also employed to adjust other risk factors. In conclusion, SBP and DBP and,?, PWV were elevated by increase of PAl1, TAT, respectively, i n the 40+ years group even after adjustment for other possible risk factors . DD had no relation to any of the indicators of arteriosclerosis. None of the coagulation-fibrinolysis parameters had any relation to brain infarctions. J Epidemiol , 2000 ; 10 : 34-41 coagulation-fibrinolysis system, arteriosclerosis , cross-sectional study


Haruhiko Sakurai
A cross-sectional observation was performed to assess the relationship between the coagulation-fibrinolysis system and the subclinical indicators of arteriosclerosis in a healthy male population.Subjects were 445 workers (18.9-49.4, Av.36.2yrs) in viscose rayon manufacturing factories in Japan.Coagulation-fibrinolysis parameters determined were D-dimer(DD) , thrombin antithrombinIII complex (TAT), tissue plasminogen activator (TPA) ,and plasminogen activator inhibitor 1 (PAI1).The following indicators of arteriosclerosis were examined; systolic and diastolic blood pressure (SBP,DBP), stiffness parameter of the carotid artery using ultrasound (3), pulse wave velocity of the aorta (PWV), and a number of lacunar infarctions from brain MRI.After age-stratification(-29,30-39,40+yrs),the subjects were classified into quartiles by coagulation-fibrinolysis parameters.
The mean values of SBP and DBP and ,3 and PWV the prevalence of brain infarctions were compared across these quartiles by means of analysis of variance, chi-square test, respectively .Multivariate analysis was also employed to adjust other risk factors.

It has been hypothesized
that the hypercoagulable state of blood is one of the risk factors of arteriosclerosis .Several clinical reports have remarked an abnormality of the coagulationfibrinolysis system in patients with arteriosclerosis 1-5) Moreover, a small number of prospective studies confirmed that abnormality of the coagulation-fibrinolysis system was related to the incidence of ischemic heart disease and stroke 6.7 .
However, few epidemiological studies have evaluated the association between the coagulation-fibrinolysis system and arteriosclerosis.
The purpose of the present study was to assess the relationship between the blood coagulation-fibrinolysis system and the subclinical indicators of arteriosclerosis in a healthy male population.

Study population
A prospective cohort study is in progress to clarify the effects of occupational exposure to carbon disulfide (CS2) on workers' health in eleven viscose rayon manufacturing factories in Japan.Results of a baseline study conducted in 1991-1992 were described elsewhere 8, 9) .In brief, a group of 834 males, consisting of 432 CS2-exposed workers and 402 referent workers, were examined, with special reference to the effects on arteriosclerosis.Those who have present illness or past histories of chronic diseases such as hypertension, diabetes mellitus, and cardiovascular disease were excluded from the study.No CS2-related effects , excepting a higher prevalence of retinal microaneurysm , were observed.Blood coagulation-fibrinolysis parameters in plasma were determined in 445 of the 834 workers.
The mean age of the subjects was 36.2 years, with an age range from 18.9 to 49.4 years.sured by one doctor using one sphygmomanometer from the right arm of each participant, each of whom was seated at the time and had rested for at least the preceding 15 minutes.The stiffness parameter of the carotid artery was calculated with adjustment for diastolic blood pressure after determination by using ultrasound (QMF-2000XA, Hayashi Electric Co.).This parameter, a sensitive index of the elasticity of the vessel, was found to increase in patients with arteriosclerosis 10).Pulse wave velocity of the aorta was measured by PWV-200 (Fukuda Electric Co.) and followed by adjustment for diastolic blood pressure.It was known that this value reflected the stiffness of the aortic wall and increased with the progress of arteriosclerosis 11).MRI of the brain was examined in a hospital near each factory by MRI equipment with 0.5T in magnetic density.To equalize the quality of the films, the radiologist offered some suggestions to the technicians for the first two to three subjects.One radiologist read all MRI films and measured a number of lacunar infarctions in the cerebrum without detailed knowledge of the individual subjects.

Possible risk factors
Information on possible risk factors was also collected, Age, height and weight (then, BMI was calculated), occupational history, and educational history were obtained from company records.Data on smoking status, alcohol consumption, and the presence of type A behavior pattern were accumulated through a self-administered questionnaire.Then, packyears and grams of ethanol per day were calculated as indices of smoking status Table 1.Characteritics of study population by age strata.a: percentage of senior high or higher t: geometric mean , geometric standard deviation &: for TG and insulin , the data was obtained from fasting blood sample and alcohol consumption, respectively.The presence of type A behavior pattern was assessed by a questionnaire form developed by Maeda (a cut-off point was set at a score of 17 or more)12).TC, HDL, triglycerides (TG), insulin were measured from the same sample as coagulation-fibrinolysis parameters , then TO HDL was calculated for analysis .Table 1 showed the characteristics of the study population by strata representing three age groups(-29, 30-39, 40+).

Statistical analysis
After age-stratification, the subjects were classified into quartiles by coagulation-fibrinolysis parameters (named group 1-4 from the lowest) within each age strata .The mean values of the outcome variables SBP, DBP, 3, and PAN were compared across these quartiles by means of analysis of variance followed by Tukey's method.The prevalence of lacunar infarctions (using a cut-off point of 10 or more) on MRI of the brain was compared by chi-square test.
The distribution of possible risk factors including age and alcohol consumption was comparable across quartiles (data not shown) except for BMI, TC/HDL, and smoking status (table 2).
To control for possible confounders, multiple regression analysis for SBP, DBP, 13, PWV and multiple logistic regression for lacunar infarctions from MRI were applied.The relationship between the blood coagulation-fibrinolysis system and subclinical indicators of arteriosclerosis was explored using two models.Risk factors included in model A were age, BMI (<20, 20-25, *25 kg/m2), smoking status (0, <10, 10-20, ?20 packyears), alcohol consumption (0, <25, 25-50, *50 g ethanol/day), occupational history (CS, exposure, no), educational history ( junior high school, senior high school or higher), presence of type A behavior pattern (yes, no), and coagulation-fibrinolysis parameters.Quartiles were adopted as coagulation-fibrinolysis parameters.In model B , TO HDL was added to the risk factors in model A. SBP was also included in the both models for ,3 and PWV.Trends of increase in outcome variables were assessed using regression coefficients of coagulation-fibrinolysis quartiles in a linear regression model .All statistical analysis was performed using the SAS package.

RESULTS
Assessed after age-stratification (-29, 30-39, 40+), SBP, DBP were significantly elevated by the increase of PAIL, TPA and PAIL, respectively in the 40+ years strata.In the 30-39 years strata, SBP were significantly decreased by increase of TAT (table 3).The relationship between SBP and TPA in the -29 years strata and in the 40+ years strata was somewhat irregular.* were significantly elevated by increase of PAI1 in the 40+ years group.There was no trend in the relationship between * and TAT in the 30-39 years strata (table 4).
PWV were significantly elevated by increase of TAT in the 40+ years group.The relationship between PWV and TPA in the 30-39 years strata was somewhat irregular (table 4).
In the 40+ years strata, even after adjustment for other possible risk factors, SBP and DBP and 3, PWV were significantly elevated by increase of PAII,TAT, respectively (table 5).Results were the same after including fasting TG and fasting insulin in the model, for the subjects fasting blood samples were collected from.
None of the coagulation-fibrinolysis parameters had any relation to lacunar infarctions in the brain, though a tendency to be elevated by increase of age strata was prevalent (table 6).Results were the same after logistic regression analysis.

DISCUSSION
It is widely known that hypertension, hyperlipidemia, and smoking are risk factors for arteriosclerosis.Recently, it has been hypothesized that abnormality of the coagulation-fibrinolysis system is an additional risk factor of arteriosclerosis.Under that hypothesis, the hypercoagulable state is supposed to be associated not only with the generation of thrombus but with arteriosclerosis itself.Recent clinical evidences has supported this hypothesis.Elevated PAI1 has been reported in deep vein thrombosis 13) and ischemic stroke 4).Elevated DD and TAT observed in stroke 6), elevated DD in coronary heart disease 14,15) and peripheral artery disease 16) have been also described.In Werner's syndrome, which is associated with accelerated atherosclerotic changes, DD, TAT, TPA, and PAI1 were all elevated 17).Whereas, there were few prospective studies.In a study of young (<45years) myocardial infarction survivors, elevated PAII levels were related to the development of recurrent myocardial infarction at three years 18.19).Ridker and coworkers demonstrated that elevated TPA levels were associated with future myocardial infarction and stroke 6,7).Jansson found that elevated TPA levels were a predictor of mortality in patients with angina pectoris 20, 21).In contrast, Hamsten and coworkers found that lower levels of TPA were significantly related to reinfarction in myocardial infarction survivors 18,19).The connection between TPA and coronary heart disease remains controversial.The end points in the above-mentioned studies were clinical events.Studies which have assessed directly the relationships between the coagulation-fibrinolysis system and indicators of arteriosclerosis in healthy subjects were few 22).Thus, the present study was undertaken to examine relationships between the coagulation-fibrinolysis system and the subclinical indicators of arteriosclerosis in comparatively young male subjects.
Our findings revealed that PAI I remained a significant variable for SBP and DBP even after adjustment (table 5).These findings were compatible with the hypothesis that hypofibri-noaysis, that is, hypercoagulable state, played a causative role in the development of arteriosclerosis.However, the possibility that increase of PAI1 in the present study might be a consequence rather than a cause of elevated blood pressure could not be denied.The explanatory mechanism that elevated blood pressure developed injury of endothelium and increase in releasing PAII from endothelium could be supposed.Although Vaziri demonstrated that DD was associated with blood pressure 23), no relationship was found in the present study.
* seemed to be elevated by increase of PAI1 even after adjustment for another possible risk factors (table 5).This finding was also consistent with the studies of Salomaa 22) .Although the PAII antigen was measured instead of the PAI I activity in the present study, it was well-known that both were strongly correlated 24,25).
TC/HDL, a known major risk factor for arteriosclerosis , could be interpreted as a confounding factor or as a series of consequence.If TC/HDL is considered a confounding factor , it should be included in the model; if it is considered a series of consequence, it should not be included.However, there was no difference in results between model A (which did not include Table 5. Relationship between the coagulation-fibrinolysis parameters and outcome variables, SBP, DBP, *, PWV in the 40+ years group.Crude mean values and adjusted mean values were compared across quartiles#.

#. TC/HDL) and model B (which included TC/HDL).
PWV, which represents the degree of arteriosclerosis in the aorta, were elevated by increase of TAT, but demonstrated no relation to DD, TPA, and PAIL.There was no discernible reason why relationship between 13 and PAI I observed in the carotid artery, whereas relationship between PWV and TAT observed in aorta.The possibility that the mechanism in the development of arteriosclerosis might differ between in the carotid artery and in aorta exists.Interpretation of the result, however, must be made with caution because of the difference in the method of measurement; pulse wave velocity was used for the aorta, and ultrasound for the carotid artery.
Few studies have demonstrated the association between the coagulation-fibrinolysis system and lacunar infarctions identi-fied from magnetic resonance imaging of the brain.A recent study showed that the number of small infarctions defined as lesions less than 10 mm in diameter present on MRI was associated significantly with the concentration of plasma fibrinogen and reversely with the activity of antithrombin U1, an inhibitory factor in coagulation system mot.These results indicated that a hypercoagulable state might causatively relate with small vessel disease in the brain, whereas, in the present study, none of the coagulation-fibrinolysis parameters had any relation to the number of lacunar infarctions on MRI of the brain.In general, lacunar infarctions observed on MRI of the brain were thought to be small vessel lesions based on 'arteriolosclerosis' rather than middlesized or large vessel lesions based on atherosclerosis.
The possibility that the coagulation-fibrinolysis sys-Table 6. Relationship between the coagulation-fibrinolysis parameters and lacunar infarctions on MRI of the brain.Prevalence of lacunar infarctions (a cut-off was set at 10 or more) was compared by chi-square test.
tem might not be significantly involved in arteriolosclerosis exists.
Potential limitations of this study should be discussed .First, the subjects were all male.Second, the limitation of the subjects' occupation existed.These mean a potential lack of general applicability.Third, measuring the antigen of TPA and PAI1 instead of the activity of each of these might, as remarked earlier, present problems.
However, consider the advantages of the present study , the first being that the subjects are comparatively young and healthy.Few studies are available in which subjects are as yet young and healthy.Secondly, there is the possibility of tracing this cohort to assess the causative relationship between the coagulation-fibrinolysis system and the development of arteriosclerosis which is observed as increase in the subclinical indicators.Lastly, it is advantageous that the targets included various types of arteries, aorta, carotid artery and small vessels in the brain.
In conclusion, SBP and DBP and 3, PWV were elevated by increase of PAT 1, TAT, respectively, in the 40+ years group even after adjustment for other possible risk factors in a healthy male population.

Table 2 .
The distribution of possible risk factors.The mean values of BMI , TC/HDL, and smoking status were compared across quartiles in each age strata by means of analysis of variance .

Table 3 .
Relationship between the coagulation-fibrinolysis parameters and SBP, DBP.The mean values of SBP, DBP were compared across quartiles in each age strata by means of analysis of variance.

Table 4 .
Relationship between the coagulation-fibrinolysis parameters and 3, PWV.The mean values of 3, PWV were compared across quartiles in each age strata by means of analysis of variance.