Segregation Analysis of Plasminogen Activator Inhibitor-1 and Fibrinogen Levels in the NHLBI Family Heart Study

From the Department of Epidemiology, School of Public Health, University of North Carolina at Chapel Hill (J.S.P.); Division of Epidemiology, School of Public Health, University of Minnesota (A.R.F., T.A.S.), and Department of Laboratory Medicine and Pathology, University of Minnesota Medical School (J.E.), Minneapolis, Minn; Division of Biostatistics, Washington University School of Medicine, St Louis, Mo (M.A.P., D.C.R.); and Cardiovascular Genetics, University of Utah, Salt Lake City, Utah (R.R.W.).

Correspondence to Dr James S. Pankow, Department of Epidemiology, School of Public Health, University of North Carolina at Chapel Hill, NationsBank Plaza, Suite 306, 137 E Franklin St, Chapel Hill, NC 27514. E mail jim_pankow@unc.edu

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Abstract—Elevated plasminogen activator inhibitor-1 (PAI-1) and fibrinogen concentrations are risk factors for coronary heart disease. We investigated environmental, familial, and genetic influences on PAI-1 antigen and fibrinogen concentrations in 2029 adults from 512 randomly ascertained families in 4 US communities. We used maximum-likelihood segregation analysis to fit several genetic and nongenetic modes of inheritance to the data to determine whether mendelian inheritance of a major gene could best explain the familial distributions of these 2 hemostatic factors. Age- and gender-adjusted familial correlations for PAI-1 antigen level averaged 0.16 in first-degree relatives (95% CI=0.11 to 0.21); the spouse correlation was positive but not statistically significant (r=0.10, 95% CI=-0.02 to 0.23). Complex segregation analysis indicated a major gene associated with higher PAI-1 concentrations in 65% of individuals from these families. Demographic, anthropometric, lifestyle, and metabolic characteristics together explained 37% to 47% of the variation in PAI-1 antigen levels, and the inferred major gene explained an additional 17% of the variance. Positive and statistically significant age- and gender-adjusted familial correlations in first-degree relatives indicated a possible heritable component influencing plasma fibrinogen concentration (r=0.17, 95% CI=0.13 to 0.22); however, segregation analysis did not provide statistical evidence of a major gene controlling fibrinogen level. These family data suggest that there are modest familial and genetic effects on the concentration of PAI-1.

Key Words: plasminogen activator inhibitor-1 • fibrinogen • heritability • segregation analysis

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Thrombosis after atherosclerotic plaque rupture plays a major role in acute myocardial infarction and sudden cardiac death. The balance of systemic thrombotic and fibrinolytic forces at the time of plaque disruption may determine the size, stability, and persistence of developing thrombi. Consistent with this hypothesis, prospective epidemiological studies have reported that elevated plasminogen activator inhibitor-1 (PAI-1) levels or elevated fibrinogen concentrations measured in middle-aged adults predict subsequent incident or recurrent coronary events.^{1 2 3 4} Elevated PAI-1 and fibrinogen may be causes of coronary heart disease (CHD), intermediates in the etiologic pathway linking traditional risk factors with CHD, or simple markers of subclinical atherosclerosis and chronic, low-grade inflammation.⁵

Although the demographic, anthropometric, lifestyle, and metabolic correlates of PAI-1 and fibrinogen have been well characterized, few population-based studies have examined the familial and genetic determinants of these hemostatic factors. We investigated familial, polygenic, and major gene effects on plasma concentrations of PAI-1 and fibrinogen in 512 randomly ascertained families using 2 statistical methods, familial correlation analysis and segregation analysis. Our data provide evidence of a major gene influencing plasma PAI-1 levels.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
The National Heart, Lung, and Blood Institute Family Heart Study (FHS) is an investigation of genetic and nongenetic determinants of CHD, preclinical atherosclerosis, and cardiovascular risk factors.⁶ Unrelated individuals (probands) were selected from ongoing population-based cohort studies in 4 US communities. In 2 of the communities (Forsyth County, NC, and suburban Minneapolis, Minn), probands were participants in the Atherosclerosis Risk in Communities Study. In Salt Lake City, Utah, probands were participants in the Utah Health Family Tree Study, and in Framingham, Mass, probands were offspring of members of the original Framingham Cohort Study. Because the Utah and Framingham studies had already recruited some biological relatives, 1 individual was selected randomly from each eligible sibship to ensure that probands for FHS were unrelated. A total of 14 592 probands were identified in the 4 communities. From lists of eligible probands, a random sample of 500 families and a nonrandom sample of 500 high-risk families were selected at each center. We included only members of randomly ascertained families in this analysis.

Family History Questionnaires (Phase I)
Probands were mailed a family history questionnaire (FH1) and asked to provide information about their parents, siblings, spouses, and children, including demographic information and history of myocardial infarction, coronary procedures, angina, stroke, diabetes, and other diseases. Information on deceased relatives was also collected. The proband and each living family member named by the proband on the FH1 questionnaire were then mailed a personal health history questionnaire (FH3). Participants were asked to provide a detailed medical history, including diagnosis of or hospitalization for myocardial infarction, coronary procedures, angina, stroke, diabetes, and other diseases. The participation rate for probands was 67%; response rates varied from 63% to 82% across centers. Approximately 86% of eligible relatives completed the FH3 questionnaire; response rates varied from 78% to 94% across centers.

Physical Examination (Phase II)
Selected probands 45 years and older and their immediate family members (parents, siblings, children, current or former spouses) 25 years and older were invited for a comprehensive physical examination at a local clinic if their family satisfied minimum participation requirements in phase I of the study. A total of 2673 individuals from 541 random families completed either a full or abbreviated examination. The study was approved by an institutional review committee at each site, and subjects gave informed consent.

Participants were asked to fast at least 12 hours before arrival at the clinic. Blood was drawn from an antecubital vein with free blood flow and minimal trauma while patients were seated. Specimens for PAI-1 and fibrinogen assays were collected in vacuum tubes containing sodium citrate. Immediately after venipuncture, samples were placed in an ice-water bath, then centrifuged at 3000g for 10 minutes at 4°C. Plasma was removed, with special care taken to withdraw only the plasma and not the buffy coat containing platelets or the lipid layer, which could adversely affect the assays on specimens. Samples were placed in a -70°C freezer no more than 90 minutes after venipuncture. Frozen samples were packaged in dry ice, shipped to the FHS central laboratory at the Department of Laboratory Medicine and Pathology, University of Minnesota, Minneapolis, and stored at -70°C until fibrinogen and PAI-1 assays could be completed.

Measurement of Hemostatic Variables
PAI-1 antigen was measured in plasma using an enzyme-linked immunosorbent assay.⁷ Mouse monoclonal anti-human PAI-1 antibodies coupled with peroxidase were directed against PAI-1 from plasma, and the amount of orthophenylenediamine substrate cleaved by the peroxidase was used to estimate PAI-1 concentration. Reagents were from the Diagnostica Stago Asserachrom PAI-1 kit 00577 (American Bioproducts). The color intensity of the sample was measured using a MR700 Microplate Reader (Dynatech Laboratories). The coefficient of variation for PAI-1 antigen at 8 and 152 ng/mL was 21.2% and 8.8%, respectively.

Fibrinogen concentration was measured in plasma using the Clauss method.⁸ The rate of conversion of fibrinogen to a fibrin clot (clotting time) was measured using excess bovine thrombin (Parke Davis) and an MLA Electra 800 Automatic Coagulation Timer (Medical Laboratory Automation, Inc). Clotting times of samples were compared with clotting times measured at various dilutions of standard plasma (Dade Data-Fi Fibrinogen Calibration Reference, Dade Baxter Scientific Products). Results using standard dilutions were plotted, and fibrinogen concentrations were estimated from the graph. The coefficient of variation for fibrinogen at 1.5 and 2.6 g/L was 6.7% and 3.8%, respectively.

Other Measurements
Current and former cigarette smoking habits were ascertained by questionnaire. Usual consumption of alcohol was estimated from self-reported weekly intake of wine, beer, and liquor. Intake in grams per week was computed by multiplying the average alcohol content of each source (wine, 10.8 g; beer, 13.2 g; liquor, 15.1 g) by the number of drinks consumed in a typical week. Average weekly physical activity during the previous year was estimated from self-reported frequency and duration of light, moderate, and strenuous exercise. In women, a reproductive history questionnaire was used to determine menopausal status and a medication inventory was used to ascertain current use of oral contraceptives and replacement hormones.

Standing height, rounded down to the nearest centimeter, was measured using a wall-mounted vertical metal ruler. Body weight was recorded to the nearest pound using a balance scale. Waist and hip circumferences, rounded to the nearest centimeter, were measured at the level of the umbilicus and at the maximum protrusion of the gluteal muscles, respectively. Body mass index (kg/m²) and waist-to-hip ratio were computed. Serum insulin was measured by using a radioimmunoassay (Coat-A-Count, Diagnostic Products Corp). Serum glucose was measured on the Kodak EKTACHEM Clinical Chemistry Slide.

The medical history of each participant was updated during the physical examination to include recent diagnoses of myocardial infarction, angina, stroke, diabetes, or other diseases. An electrocardiogram was obtained to assess known and silent Q-wave myocardial infarction, ventricular hypertrophy, ischemia, and other indicators of cardiac function. We defined prevalent cardiovascular disease as self-reported personal history of myocardial infarction, coronary angioplasty, coronary artery bypass surgery, stroke, angina, or electrocardiographic evidence of myocardial infarction (major Q-wave elevation [Minnesota codes 1.1 or 1.2] or minor Q-wave and ST elevation [Minnesota code 1.3 and codes 5.1 or 5.2]). We defined diabetes as self-reported history of diabetes, nonfasting glucose level of 200 mg/dL, fasting glucose level of 140 mg/dL, or current pharmacological treatment for diabetes.

Statistical Methods
Familial Correlations
We used the SEGPATH program⁹ to estimate familial correlations for PAI-1 antigen level and fibrinogen concentration. SEGPATH is a general purpose program, based on linear path models, that can estimate gender-specific familial correlations using maximum-likelihood methods. Before correlation analysis, PAI-1 and fibrinogen levels were standardized to a mean of 0 and an SD of 1.

In SEGPATH, we fitted a general model to the data to simultaneously estimate all 8 nuclear family correlations (mother-father, father-son, father-daughter, mother-son, mother-daughter, son-son, daughter-daughter, and son-daughter). We also fitted more restrictive models to the data (eg, assuming equal correlations for all parent-offspring pairs and all sibling pairs). We estimated maximal heritability by doubling the age- and gender-adjusted correlation estimate for first-degree relatives.¹⁰

Segregation Analysis
We conducted complex segregation analysis of PAI-1 antigen level and fibrinogen concentration using class D regressive models¹¹ as implemented in the REGC program.¹² We fitted a series of explicit genetic and arbitrary nongenetic modes of inheritance to the data and determined whether they provided an adequate fit to the data compared with a general model that was not constrained by mendelian patterns of inheritance. To compare the relative fit of each restricted model to the general model, we performed likelihood ratio tests by computing twice the difference in log likelihood values between the 2 models and comparing the result to a ² distribution, with the number of degrees of freedom equal to the difference in the number of estimated parameters between models. Under this approach, more restrictive genetic or nongenetic models that provide a relatively poor fit to the data will have a high ² value and will be rejected in favor of the general model. We compared nonnested models using Akaike's Information Criterion (AIC).¹³ The AIC [-2 log likelihood+2(no. of estimated parameters)] incorporates a penalty for models fitting extra parameters; the more parsimonious model will have a lower AIC value.

We assumed that a major gene effect on the 2 phenotypes (PAI-1 or fibrinogen), if present, was due to an unmeasured autosomal locus with 2 alleles (designated H and L) associated with higher (H) and lower (L) values of the phenotype. Three genotypes are possible under this single-locus model (HH, HL, and LL). When mendelian inheritance is not established, a more general term, "ousiotype," may be used to describe individuals belonging to a distinct phenotypic subgroup or type.¹⁴

For mendelian modes of inheritance, basic parameters included the frequency of the allele associated with the higher value of the phenotype (q_H), the mean of the phenotype for each specific type (µ_HH, µ_HL, and µ_LL), and the residual variance of the phenotype within each type (²). The transmission parameters (_HH, _HL, and _LL) define the probability that a parent of a specific type transmits an "H" allele to his or her offspring. For mendelian modes of inheritance, these probabilities are fixed at 1.0, 0.5, and 0.0 for the HH, HL, and LL types, respectively. For the general or unrestricted model, these transmission probabilities are estimated rather than fixed; deviations of _HH, _HL, or _LL from mendelian expectations may be due to a number of factors, including the presence of several major genes or genotype-environment interactions. We included residual spouse (_SP), mother-offspring (_MO), father-offspring (_FO), and sibling (_SS) correlation parameters in several of our models to account for sources of familial resemblance other than a major gene, such as polygenic effects, shared environmental effects, or cultural transmission. We assumed that the distribution of types was in Hardy-Weinberg equilibrium and that the residual variance of the phenotype was the same for all types (ie, ²_HH=²_HL=²_LL).

We inferred that a phenotype (PAI-1 or fibrinogen) was under the control of a major gene if the results of segregation analysis satisfied 3 widely accepted statistical criteria¹⁵ : (1) rejection of the hypothesis of no major effect (q_H=1), (2) rejection of the hypothesis of no transmission of the major effect (_HH=_HL=_LL), and (3) nonrejection of the hypothesis of mendelian transmission (_HH=1, _HL=0.5, and _LL=0).

Covariate Adjustments
In preliminary analyses, we found that the statistical association between PAI-1 or fibrinogen and several covariates was modified by gender. Because interaction effects (eg, the interaction of gender and age) cannot be accommodated in segregation analysis using the REGC program, we used a 2-stage approach to account for covariates.¹⁶ In the first stage, we used multiple linear regression to adjust the phenotype (PAI-1 or fibrinogen) for the effect of covariates. In the second stage, we fitted selected modes of inheritance to the data using the adjusted phenotype values obtained in the first stage.

In the regression analysis (first stage), we adjusted the phenotypes on groups of related variables. These groups of variables included age (age and age squared), anthropometric characteristics (height, weight, and waist-to-hip ratio), lifestyle characteristics (smoking status, current cigarettes smoked per day, drinking status, current alcohol intake, physical activity, oral contraceptive use, and hormone replacement therapy), and metabolic characteristics (diabetes and fasting serum insulin). We fitted regression models separately for women and men and included covariates in these models regardless of their statistical significance. Before conducting segregation analysis (second stage), we added a constant reflecting the sample mean [(ln)PAI-1, 2.58; fibrinogen, 3.08 g/L] to each individual's residual phenotype value.

Exclusions
Although families were independently ascertained, some participants were members of more than 1 extended pedigree (n=61). We duplicated phenotype and covariate data for these participants to retain individuals in multiple pedigrees and to maximize the amount of information available for genetic analyses. We excluded members of 20 families because of insufficient information about the family structure. We randomly excluded 1 member of each identical twin pair (n=7) because the genetic epidemiological programs used here do not appropriately model identical twin relationships. Only 2381 of the 2706 remaining participants (88%) had PAI-1 or fibrinogen assays, in part because some participants completed an abbreviated examination that did not permit collection of specimens for coagulation studies. We excluded 235 individuals with prevalent cardiovascular disease because PAI-1 and fibrinogen are acute-phase reactants that may rise nonspecifically during acute and chronic inflammatory episodes, such as those associated with atherosclerosis.¹⁷ No participants were excluded on the basis of measures of subclinical atherosclerosis, such as carotid intimal-medial thickness or the ankle-brachial systolic pressure index. For the 133 participants who reported a history of myocardial infarction, angioplasty, or coronary bypass surgery, 101 events (76%) were validated by review of existing medical records; however, no data were obtained to confirm self-reported angina or stroke. We also excluded participants missing medical information or electrocardiographic results required to assess the presence or absence of cardiovascular disease (n=90) and individuals who took anticoagulants (coumarin or heparin) within 2 weeks of the examination (n=27). After all exclusions, the sample included 2029 individuals from 512 randomly ascertained families. Including probands, there were 252 families with 1 to 3 members, 187 families with 4 to 6 members, 59 families with 7 to 9 members, and 14 families with 10 to 12 members. We excluded 3 participants with adjusted fibrinogen values 7.4 g/L from segregation analysis to prevent their extreme phenotype values from unduly influencing the parameter estimates.

	Results

Top Abstract Introduction Methods Results Discussion References

 
Participant Characteristics
Participants ranged in age from 25 to 91 years; the median age was 52 years. Nearly 60% of the women were postmenopausal, and 30% were currently taking replacement hormones (Table 1). The prevalence of smoking (11% to 14%) was about 50% lower than estimates from surveys of the US adult population.¹⁸ The prevalence of diabetes was 5% in both women and men.

Plasma PAI-1 antigen levels ranged from 0.2 to 300 ng/mL, with a median of 13 ng/mL. The distribution of PAI-1 values exhibited extreme positive skewness (coefficient of skewness, 2.59). Because skewness can lead to false acceptance of a major gene in segregation analysis,¹⁹ we normalized the PAI-1 distribution using the natural logarithmic transformation (coefficient of skewness after transformation, -0.02). Geometric mean values of PAI-1 were higher in men than women (17.8 versus 10.8 ng/mL). Fibrinogen concentrations ranged from 1.5 to 6.1 g/L in the entire sample, and the distribution was slightly positively skewed. Mean levels of fibrinogen were 3.18 g/L in women and 2.97 g/L in men (Table 1).

Correlates of PAI-1
In women, PAI-1 antigen level was positively associated with age, weight, waist-to-hip ratio, diabetes, and fasting serum insulin level and negatively associated with height and physical activity. PAI-1 was lower in women currently using oral contraceptives or replacement hormones. In men, PAI-1 level was negatively associated with age and positively associated with weight, waist-to-hip ratio, alcohol intake, diabetes, and fasting serum insulin level. Demographic, anthropometric, lifestyle, and metabolic characteristics together explained 47% and 37% of the variance of PAI-1 concentration in women and men, respectively (Table 2). Among both women and men, the largest incremental change in the model R² was attributable to anthropometric variables (weight, height, and waist-to-hip ratio).

Familial Correlations for PAI-1
Age- and gender-adjusted familial correlations for PAI-1 antigen level ranged from 0.09 (mother-son) to 0.29 (sisters), with an average of 0.16 (95% CI=0.11 to 0.21) among all first-degree relatives (Table 3). In contrast, the age- and gender-adjusted spouse correlation was not statistically significantly different from 0 (r=0.10, 95% CI=-0.02 to 0.23). Excluding spouse pairs, the correlation estimate for same-gender relatives (r=0.22) was significantly different from that for opposite-gender relatives (r=0.12) (P<0.005). Some, but not all, correlation estimates were reduced on further adjustment for anthropometric, lifestyle, and metabolic factors, which may also aggregate within families (Table 3).

Segregation Analysis of PAI-1
We first conducted segregation analysis on PAI-1 values adjusted for age and gender; the results did not provide statistical evidence of a major gene influencing plasma PAI-1 concentration (results not shown). We then repeated segregation analysis after further adjusting PAI-1 values for anthropometric characteristics (weight, height, and waist-to-hip ratio). The results are summarized in Table 4. We provide parameter estimates, model fit, and test statistics for 4 hypothetical modes of inheritance: (1) sporadic, which allows random environmental effects but no genetic transmission of the phenotype; (2) familial correlations, which allows familial resemblance for the phenotype but no major gene effect; (3a–3c) mendelian, which allows a major gene effect as well as other sources of familial resemblance; and (4) no transmission, which allows a major effect (3 distinct phenotypic subgroups or "types") but no transmission of the major effect from parents to offspring.

We assessed the goodness of fit for each of the restricted models (1 to 4) by comparing them to the general or unrestricted model (5). The sporadic and familial correlations models were both rejected (P<0.001), as was the model of no transmission of the major effect (P<0.001). In contrast, neither the mendelian-codominant nor the mendelian-dominant models with familial correlations were rejected (P0.08), suggesting that both of these models provide an adequate fit to the family data. The mendelian-dominant model had the lowest AIC value and was the most parsimonious of all the models tested. The maximum-likelihood estimates for the transmission parameters under the general model (_HH=1, _HL=0.45, and _LL=0) closely resembled the probabilities expected if there is mendelian transmission of a major gene effect. When we fitted similar genetic and nongenetic modes of inheritance to the family data after further adjusting PAI-1 values for lifestyle and metabolic variables, parameter estimates were slightly different than those presented in Table 4, but the main findings of segregation analysis were qualitatively unchanged. Together these findings meet the statistical criteria required to infer the presence of a major gene influencing PAI-1 level.

We used maximum-likelihood estimates from the best-fitting model (mendelian dominant) to estimate the proportion of the variance in PAI-1 antigen level that can be explained by the putative major gene. Regression analyses indicated that the variance of (ln)PAI-1 level was reduced from 1.71 to 1.07 (37%) when we controlled for age, weight, height, and waist-to-hip ratio. The variance of adjusted (ln)PAI-1 was further reduced to 0.78 when a major gene effect was included (model 3b, Table 4). These results suggest that the putative major gene explains 17% of the total variance of (ln)PAI-1, ie, (1.07-0.78)/1.71, and 27% of the remaining variance if PAI-1 is first adjusted for age and anthropometric variables, ie, (1.07-0.78)/1.07.

Figure 1 shows predicted genotype-specific distributions for (ln)PAI-1 antigen level based on maximum-likelihood parameter estimates from the mendelian-dominant model with familial correlations (Table 4). Genotype-specific curves are superimposed on the actual distribution of adjusted (ln)PAI-1 (shown as a histogram). According to the allele frequencies obtained under the mendelian-dominant model (q_H), we estimate that 65% of the population carry 1 or 2 alleles associated with higher levels of PAI-1 (HH or HL). The remaining 35% are homozygous for the allele associated with lower values (LL). Geometric means for these putative high and low genotypes are 19.7 and 6.8 ng/mL, respectively.

View larger version (32K): [in this window] [in a new window]   Figure 1. Histogram showing adjusted PAI-1 levels among 1989 members of 512 families from the FHS. The 3 normal distributions show the predicted distribution for the LL genotype (solid black curve), the predicted distribution for the HL and HH genotypes (solid gray curve), and the sum of the 2 predicted distributions (dashed curve). Estimated means and the common SD of PAI-1 antigen from the mendelian-dominant model (model 3b, Table 4) were used to derive the predicted distributions.

Correlates of Fibrinogen
In women, plasma fibrinogen concentration was positively associated with age, weight, and number of cigarettes smoked and negatively associated with height and alcohol intake; the mean fibrinogen level was lower in women currently taking replacement hormones. In men, fibrinogen level was positively associated with age, weight, waist-to-hip ratio, and number of cigarettes smoked and negatively associated with height and alcohol intake. Age and anthropometric, lifestyle, and metabolic characteristics together explained 19% and 29% of the variance of fibrinogen concentration in women and men, respectively (Table 2).

Familial Correlations for Fibrinogen
Age- and gender-adjusted correlations among relative pairs ranged from 0.11 (father-son) to 0.21 (mother-son and mother-daughter) (Table 3). The age- and gender-adjusted correlation coefficient for all first-degree relatives was positive and statistically significant (r=0.17, 95% CI=0.13 to 0.22); in contrast, the confidence interval for the spouse correlation included 0 (r=0.12, 95% CI=0.00 to 0.23). Further adjustments for anthropometric, lifestyle, and metabolic characteristics attenuated many of the familial correlations, but not uniformly so (Table 3).

Segregation Analysis of Fibrinogen
Table 5 presents results of segregation analysis of fibrinogen level adjusted for age, gender, and anthropometric characteristics. When compared with the general model, the sporadic, familial correlations, and mendelian-codominant models provided a poor fit to the family data and were rejected (P<0.001). However, the no transmission model (equal values), which allows a mixture of phenotypic subgroups but no mendelian transmission of a major gene effect from parents to offspring, was not rejected (P=0.34). According to the AIC, the no transmission model was the most parsimonious of the all the models tested. The key findings of segregation analysis (ie, rejection of all mendelian hypotheses and nonrejection of the hypothesis of no transmission of the major effect) were the same when fibrinogen values were further adjusted for lifestyle and metabolic characteristics (results not shown).

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
PAI-1 is a major determinant of fibrinolytic activity.²⁰ Plasma PAI-1 activity level and antigen concentration are risk factors for coronary events in patients with clinically recognized cardiovascular disease, but longitudinal data in healthy populations are lacking. Elevated PAI-1 may be part of the etiologic pathway linking insulin resistance, atherosclerosis, and its clinical manifestations²¹ ; our study confirms the strong association reported elsewhere^{22 23} between PAI-1 concentration and components of the multiple metabolic syndrome, including excess weight, abdominal obesity, diabetes mellitus, and serum insulin concentration. Increased number, mass, or biochemical activity of adipose tissue may account for elevated concentrations of circulating PAI-1.^{24 25 26}

In our study, the average age- and gender-adjusted correlation for plasma PAI-1 among first-degree relatives was 0.16, suggesting a maximum heritability of 32%. In contrast, the spouse correlation of 0.10 was positive but not statistically significant. Familial correlations were higher in same-gender than in opposite-gender relative pairs, possibly because of gender differences in endogenous hormone levels.²⁷ Several studies previously estimated the degree to which plasma PAI-1 level aggregates within families. Hong et al²⁸ reported a heritability of 42% for PAI-1 concentration in a study of 217 middle-aged and elderly twin pairs from Sweden. In a study of white European nuclear families, Henry et al²⁹ reported age- and gender-adjusted familial correlations for PAI-1 antigen level ranging from 0.21 to 0.23 in first-degree relatives but even higher correlations in spouse pairs (r=0.31).

Our results from segregation analysis provide statistical evidence that a major gene regulates PAI-1 antigen level. We detected this major gene only after removing the effects of height, weight, and waist-to-hip ratio, suggesting that genetic influences on PAI-1 are modest compared with the relatively strong effects of obesity and abdominal adiposity. Anthropometric factors together explain almost one third of the variance of plasma PAI-1 level in our study, whereas the putative major gene explains just 17% of the variance.

The PAI-1 structural locus on chromosome 7 may represent the major gene regulating PAI-1 concentration. A common deletion/insertion polymorphism (4G/5G) has been identified in the 5' promoter region; in vitro, there is a 5- to 6-fold greater responsiveness of the deletion allele (4G) to stimulation by interleukin-1,³⁰ possibly because of differential binding of a repressor protein.³¹ One small cross-sectional study indicated that individuals with 2 copies of the 4G allele had a 2-fold greater prevalence of myocardial infarction, but this finding has not been replicated in large-scale cross-sectional³² or prospective studies.³³ Several cross-sectional studies have also reported a positive correlation between the number of 4G alleles and plasma PAI-1 activity or antigen levels.^{30 31 32 34 35} However, the effect of the 4G/5G polymorphism on PAI-1 levels is not particularly strong. A recent study found that the polymorphism explained <5% of the interindividual variation in PAI-1 antigen levels in healthy individuals after adjusting for components of the multiple metabolic syndrome.²⁹

Fibrinogen is now widely recognized as an independent risk factor for CHD. An elevated plasma fibrinogen concentration may promote platelet aggregation by binding to platelet glycoprotein IIb/IIIa receptors, limit perfusion of the coronary arteries by increasing whole-blood viscosity, or accelerate the progression of atherosclerosis by infiltrating the vessel wall.^{36 37 38} In a meta-analysis of 6 cohort studies, Ernst and Resch estimated that individuals in the upper tertile of fibrinogen concentration had a 2.3-fold greater risk of subsequent cardiovascular disease than individuals in the lower tertile.

The cross-sectional correlates of plasma fibrinogen concentration include age, obesity, cigarette smoking, and alcohol intake.^{40 41 42 43 44} Our study found that 19% to 29% of the variance in plasma fibrinogen concentration could be explained by age and anthropometric, lifestyle, and metabolic factors. The extent to which genetic factors determine the concentration of plasma fibrinogen is not entirely known. Heritability estimates range from 27% to 51% in twin and pedigree studies.^{45 46 47 48 49} In our study, the average age- and gender-adjusted correlation for first-degree relative pairs suggests a maximum heritability of 34% for fibrinogen concentration. The age- and gender-adjusted spouse correlation (0.12) was not statistically significantly different from 0. Nevertheless, one might expect members of spouse pairs to be mutually exposed to acute determinants of fibrinogen, such as bacterial infections or psychological stress, often shared by members of the same household but not necessarily shared by adult family members living apart (ie, parents and offspring or siblings).

In a segregation analysis of 204 3-generation families, Livshits et al⁴⁹ found evidence for a major codominant gene effect on fibrinogen level. Our results are more in line with those of Friedlander et al,⁴⁸ who found that their data on 82 pedigrees were not consistent with mendelian transmission of a major gene regulating fibrinogen concentration. The no transmission model may have provided the best fit to the family data in our study because of residual skewness in the fibrinogen distribution, a real mixture of heterogeneous phenotypic subgroups in the study population, or a common, unmeasured environmental factor with large effects on plasma fibrinogen concentration. Genotype-environment interactions are also known to distort transmission patterns from their mendelian expectations.⁵⁰ Several observational studies have reported that the association between plasma fibrinogen concentration and cigarette smoking^{51 52 53} or strenuous physical activity⁵⁴ is influenced by a polymorphism in the promoter of the ß-fibrinogen gene. If similar genotype-environment interactions were influential in our study population, then the parameter estimates from segregation analysis may have been biased and power to infer a major gene effect may have been reduced.⁵⁵

The 3 fibrinogen structural genes (FGA, FGB, and FGG) on chromosome 4 may account for some interindividual differences in plasma fibrinogen concentration. Some, but not all, studies have reported statistically significant associations between fibrinogen level and 1 or more polymorphisms in these 3 genes.^{46 51 52 53 56 57 58 59 60 61 62 63 64 65} One large-scale population-based study that enrolled >1200 adult subjects from Ireland and France found that fibrinogen gene haplotypes explained only 1% to 2% of the variance of fibrinogen concentration. These equivocal data may indicate that other genes play a role in the regulation of plasma fibrinogen level.

With >2000 individuals from 512 random families, the present study had excellent power to estimate familial correlations and to investigate major gene influences on PAI-1 and fibrinogen level. However, several factors may have effectively reduced the power to detect a major gene. Although we excluded participants with prevalent cardiovascular disease to reduce the possible confounding effects of disease on plasma levels of the hemostatic factors, some individuals in our sample may have had elevated PAI-1 and fibrinogen levels due to chronic, low-grade inflammation associated with subclinical atherosclerosis.⁶⁶ Furthermore, average family sizes were small and many families included only 2-generation nuclear families, which are less informative than larger pedigrees with 3 or more generations. Finally, there is considerable error associated with single measurements of these hemostatic factors; the within-person (biological) and method variability for either trait is as high as 28%.^{67 68} Because residual environmental variance inflates the total variance of these 2 hemostatic factors, true genetic effects may have been underestimated in familial correlation analysis and segregation analysis because only 1 blood sample was available to estimate the habitual level for each participant.

Segregation analysis is practically constrained by the limited number of genetic and nongenetic modes of inheritance that can be fitted to the data. In our analysis, we assumed a single major gene with 2 alleles. Our relatively simple mendelian single-locus models may have been inadequate if more complex patterns of inheritance, such as oligogenic effects, multiple alleles, epistasis, and gene-environment interactions explain the distribution of hemostatic factors within families. Although we found statistical evidence of a major gene influencing plasma PAI-1 concentration, our evidence remains circumstantial until it is confirmed by linkage analysis or other molecular genetic studies.

In conclusion, our results are consistent with modest familial and genetic influences on the plasma concentrations of PAI-1 antigen and fibrinogen in randomly ascertained families from 4 US communities. We detected a putative major gene regulating PAI-1 levels and explaining 17% of the overall variance of the trait. Additional studies are needed to identify the genetic factors, environmental factors, and gene-environment interactions regulating PAI-1 and fibrinogen concentrations.

	Acknowledgments

 
We thank the FHS study participants and staff. Support was provided by National Heart, Lung, and Blood Institute cooperative agreement grants N01-HC-25104, N01-HC-25105, N01-HC-25106, N01-HC-25107, N01-HC-25108, and N01-HC-25109 and institutional training grant T32 HL07036. Some of the results presented in this report were obtained by using the program S.A.G.E., which is supported by a US Public Health Service Resource Grant (1 P41 RR03655) from the National Center for Research Resources.

Received July 24, 1997; accepted April 7, 1998.

	References

Top Abstract Introduction Methods Results Discussion References

 
1. Meade TW, Mellows S, Brozovic M, Miller GJ, Chakrabarti RR, North WRS, Haines AP, Stirling Y, Imeson JD, Thompson SG. Haemostatic function and ischaemic heart disease: principal results of the Northwick Park Heart Study. Lancet. 1986;2:533–537.[Medline] [Order article via Infotrieve]

2. Heinrich J, Balleisen L, Schulte H, Assman 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.[Abstract/Free Full Text]

3. Hamsten A, De Faire U, Walldius G, Dahlen G, Szamosi A, Landou C, Blomback M, Wiman B. Plasminogen activator inhibitor in plasma: risk factor for recurrent myocardial infarction. Lancet. 1987;2:3–9.[Medline] [Order article via Infotrieve]

4. Juhan-Vague I, Pyke SDM, Alessi MC, Jespersen J, Haverkate F, Thompson SG. Fibrinolytic factors and the risk of myocardial infarction or sudden death in patients with angina pectoris. Circulation. 1996;94:2057–2063.[Abstract/Free Full Text]

5. Tracy RP, Bovill EG. Thrombosis and cardiovascular risk in the elderly. Arch Pathol Lab Med. 1992;116:1307–1312.[Medline] [Order article via Infotrieve]

6. Higgins M, Province M, Heiss G, Eckfeldt J, Ellison C, Folsom A, Rao DC, Sprafka M, Williams R, for the NHLBI Family Heart Study Investigators. The NHLBI Family Heart Study: objectives and design. Am J Epidemiol. 1996;143:1219–1228.[Abstract/Free Full Text]

7. de Clerck PJ, Alessim C, Verstreken M, Kruithof EK, Juhan-Vague I, Collen D. Measurement of plasminogen activator inhibitor 1 in biologic fluids with a murine monoclonal antibody-based enzyme-linked immunosorbent assay. Blood. 1988;71:220–225.[Abstract/Free Full Text]

8. Clauss A. Gerinnungsphysiologische Schnellmethode zur bestimmung des Fibrinogens. Acta Haematol. 1957;17:237–246.[Medline] [Order article via Infotrieve]

9. Province MA, Rao DC. General purpose model and a computer program for combined segregation and path analysis (SEGPATH): automatically creating computer programs from symbolic language model specifications. Genet Epidemiol. 1995;12:203–219.[Medline] [Order article via Infotrieve]

10. Falconer DS. Introduction to Quantitative Genetics. 3rd ed. New York, NY: Wiley; 1989.

11. Bonney GE. Regressive logistic models for familial disease and other binary traits. Biometrics. 1986;42:611–625.[Medline] [Order article via Infotrieve]

12. Statistical Analysis for Genetic Epidemiology, release 2.2. Cleveland, Ohio: Department Epidemiology and of Biostatistics, Case Western Reserve University; 1994.

13. Akaike H. A new look at the statistical model identification. IEEE Trans Autom Control. 1974;19:716–726.

14. Cannings E, Thompson EA, Skolnick MH. Probability functions on complex pedigrees. Adv Appl Probl. 1978;10:26–61.

15. Lalouel JM, Rao DC, Morton NE, Elston RC. A unified model for complex segregation analysis. Am J Hum Genet. 1983;35:816–826.[Medline] [Order article via Infotrieve]

16. Beaty TH. Evolving methods in genetic epidemiology, I: analysis of genetic and environmental factors in family studies. Epidemiol Rev. 1997;19:14–23.[Free Full Text]

17. Stuart J, George AJ, Davies AJ, Aukland A, Hurlow RA. Haematological stress syndrome in atherosclerosis. J Clin Pathol. 1981;34:464–467.[Abstract/Free Full Text]

18. Centers for Disease Control. Cigarette smoking among young adults: United States, 1990. MMWR Morb Mortal Wkly Rep. 1992;41:354–362.[Medline] [Order article via Infotrieve]

19. Demenais F, Lathrop M, Lalouel JM. Robustness and power of the unified model in the analysis of quantitative measurements. Am J Hum Genet. 1986;38:228–234.[Medline] [Order article via Infotrieve]

20. Juhan-Vague I, Alesssi MC. Plasminogen activator inhibitor 1 and atherothrombosis. Thromb Haemost. 1993;70:138–143.[Medline] [Order article via Infotrieve]

21. Juhan-Vague I, Alessi MC, Vague P. Increased plasma plasminogen activator inhibitor-1 levels: a possible link between insulin resistance and atherothrombosis. Diabetologia. 1991;34:457–462.[Medline] [Order article via Infotrieve]

22. Potter van Loon BJ, Cluft C, Radder JK, Blankenstein MA, Meinders AE. The cardiovascular risk factor plasminogen activator inhibitor type 1 is related to insulin resistance. Metabolism. 1993;8:945–949.

23. Meilahn EN, Cauley JA, Tracy RP, Macy EO, Gutai JP, Kuller LH. Association of sex hormones and adiposity with plasma levels of fibrinogen and PAI-1 in postmenopausal women. Am J Epidemiol. 1996;143:159–166.[Abstract/Free Full Text]

24. Lundgren CH, Brown SL, Nordt TK, Sobel BE, Fujii S. Elaboration of type-1 plasminogen activator inhibitor from adipocytes: a potential pathogenetic link between obesity and cardiovascular disease. Circulation. 1996;93:106–110.[Abstract/Free Full Text]

25. Shimomura I, Funahashi T, Takahashi M, Maeda K, Kotani K, Nakamura T, Yamashita S, Miura M, Fukuda Y, Takemura K, Tokunaga K, Matsuzawa Y. Enhanced expression of PAI-1 in visceral fat: possible contributor to vascular disease in obesity. Nat Med. 1996;2:800–803.[Medline] [Order article via Infotrieve]

26. Loskutoff DJ, Samad F. The adipocyte and hemostatic balance in obesity. Arterioscler Thromb Vasc Biol. 1998;18:1–6.[Free Full Text]

27. Gebara OCE, Mittleman MA, Sutherland P, Lipinska I, Matheney T, Xu P, Welty FK, Wilson PWF, Levy D, Muller JE, Tofler GH. Association between increased estrogen status and increased fibrinolytic potential in the Framingham Offspring Study. Circulation. 1995;91:1952–1958.[Abstract/Free Full Text]

28. Hong Y, Pedersen NL, Egberg N, de Faire U. Moderate genetic influences on plasma levels of plasminogen activator-1 and evidence of genetic and environmental influences shared by PAI-1, triglycerides, and body mass index. Arterioscler Thromb Vasc Biol. 1997;17:2776–2782.[Abstract/Free Full Text]

29. Henry M, Tregouët DA, Alessi MC, Aillaud MF, Visvikis S, Siest G, Tiret L, Juhan-Vague I. Metabolic determinants are much more important than genetic polymorphisms in determining the PAI-1 activity and antigen plasma concentrations. Arterioscler Thromb Vasc Biol. 1998;18:84–91.[Abstract/Free Full Text]

30. Dawson SJ, Wiman B, Hamsten A, Green F, Humphries S, Henney AM. The two allele sequences of a common polymorphism in the promoter of the plasminogen activator inhibitor-1 (PAI-1) gene respond differently to interleukin-1 in HepG2 cells. J Biol Chem. 1993;268:10739–10745.[Abstract/Free Full Text]

31. Eriksson P, Kallin B, van `t Hooft FM, Bavenholm P, Hamsten A. Allele-specific increase in basal transcription of the plasminogen-activator inhibitor 1 gene is associated with myocardial infarction. Proc Natl Acad Sci U S A. 1995;92:1851–1855.[Abstract/Free Full Text]

32. Ye S, Green FR, Scarabin PY, Nicaud V, Bara L, Dawson SJ, Humphries SE, Evans A, Luc G, Cambou JP, Arveiler D, Henney AM, Cambien F. The 4G/5G genetic polymorphism in the promoter of the plasminogen activator inhibitor-1 (PAI-1) gene is associated with differences in plasma PAI-1 activity but not risk of myocardial infarction in the ECTIM study. Thromb Haemost. 1995;74:837–841.[Medline] [Order article via Infotrieve]

33. Ridker PM, Hennekens CH, Lindpaintner K, Stampfer MJ, Miletich JP. Arterial and venous thrombosis is not associated with the 4G/5G polymorphism in the promoter of the plasminogen activator inhibitor gene in a large cohort of US men. Circulation. 1997;95:59–62.[Abstract/Free Full Text]

34. Panahloo A, Mohamed-Ali V, Lane A, Green F, Humphries SE, Yudkin JS. Determinants of plasminogen activator inhibitor 1 activity in treated NIDDM and its relation to a polymorphism in the plasminogen activator inhibitor 1 gene. Diabetes. 1995;44:37–42.[Abstract]

35. Oessei-Gerning N, Mansfield MW, Stickland MH, Wilson IJ, Grant PJ. Plasminogen activator inhibitor-1 promoter 4G/5G genotype and plasma levels in relation to a history of myocardial infarction in patients characterized by coronary angiography. Arterioscler Thromb Vasc Biol. 1997;17:33–37.[Abstract/Free Full Text]

36. Lowe GDO, Reavey MM, Johnston RV, Forbes CD, Prentice CRM. Increased platelet aggregates in vascular and non-vascular illness: correlation with plasma fibrinogen and effect of ancrod. Thromb Res. 1979;14:377–386.[Medline] [Order article via Infotrieve]

37. Meade TW, Vickers MV, Thompson SG, Stirling Y, Haines AP, Miller GJ. Epidemiological characteristics of platelet aggregability. BMJ. 1985;290:428–432.

38. Bini A, Kudryk BJ. Fibrinogen in human atherosclerosis. Ann N Y Acad Sci. 1995;748:461–473.[Medline] [Order article via Infotrieve]

39. Ernst E, Resch KL. Fibrinogen as a cardiovascular risk factor: a meta-analysis and review of the literature. Ann Intern Med. 1993;118:956–963.[Abstract/Free Full Text]

40. Balleisen L, Bailey J, Epping P-H, Schulte H, van de Loo J. Epidemiological study on factor VII, factor VIII, and fibrinogen in an industrial population. I: baseline data on the relation to age, gender, body-weight, smoking, alcohol, pill-using, and menopause. Thromb Haemost. 1985;54:475–479.[Medline] [Order article via Infotrieve]

41. Folsom AR, Wu KK, Davis CE, Conlan MG, Sorlie PD, Szklo M. Population correlates of plasma fibrinogen and factor VII, putative cardiovascular risk factors. Atherosclerosis. 1991;91:191–205.[Medline] [Order article via Infotrieve]

42. Krobot K, Hense HW, Cremer P, Eberle E, Keil U. Determinants of plasma fibrinogen: relation to body weight, waist-to-hip ratio, smoking, alcohol, age, and sex. Arterioscler Thromb. 1992;12:780–788.[Abstract/Free Full Text]

43. Folsom AR, Qamhieh HT, Flack JM, Hilner JE, Liu K, Howard BV, Tracy RP. Plasma fibrinogen: levels and correlates in young adults: the Coronary Artery Risk Development in Young Adults (CARDIA) study. Am J Epidemiol. 1993;138:1023–1036.[Abstract/Free Full Text]

44. Cushman M, Yanez D, Psaty BM, Fried LP, Heiss G, Lee M, Polak JF, Savage PJ, Tracy RP, for the Cardiovascular Health Study Investigators. Association of fibrinogen and coagulation factors VII and VIII with cardiovascular risk factors in the elderly. Am J Epidemiol. 1996;143:665–676.[Abstract/Free Full Text]

45. Hamsten A, Iselius L, de Faire U, Blomback M. Genetic and cultural inheritance of plasma fibrinogen concentration. Lancet. 1987;2:988–991.[Medline] [Order article via Infotrieve]

46. Berg K, Kierulf P. DNA polymorphisms at fibrinogen loci and plasma fibrinogen concentration. Clin Genet. 1989;36:229–235.[Medline] [Order article via Infotrieve]

47. Reed T, Tracy RP, Fabsitz RR. Minimal genetic influences on plasma fibrinogen level in adult males in the NHBLI twin study. Clin Genet. 1994;45:71–77.[Medline] [Order article via Infotrieve]

48. Friedlander Y, Elkana Y, Sinnreich R, Kark JD. Genetic and environmental sources of fibrinogen variability in Israeli families: the Kibbutzim Family Study. Am J Hum Genet. 1995;56:1194–1206.[Medline] [Order article via Infotrieve]

49. Livshits G, Schettler G, Graff E, Blettner M, Wahrendorf J, Brunner D. Tel-Aviv–Heidelberg Three-Generation Offspring Study: genetic determinants of plasma fibrinogen level. Am J Med Genet. 1996;63:509–517.[Medline] [Order article via Infotrieve]

50. Borecki IB, Bonney GE, Rice T, Bouchard C, Rao DC. Influence of genotype-dependent effects of covariates on the outcome of segregation analysis of the body mass index. Am J Hum Genet. 1993;53:676–687.[Medline] [Order article via Infotrieve]

51. Green F, Hamsten A, Blomback M, Humphries S. The role of the ß-fibrinogen genotype in determining plasma fibrinogen levels in young survivors of myocardial infarction and healthy controls from Sweden. Haemostasis. 1993;70:915–920.

52. Humphries SE, Ye S, Talmud P, Bara L, Wilhelmsen L, Tiret L. European Atherosclerosis Research Study: genotype at the fibrinogen locus (G-455-A ß gene) is associated with differences in plasma fibrinogen levels in young men and women from different regions in Europe: evidence for gender-genotype-environment interaction. Arterioscler Thromb Vasc Biol. 1995;15:96–104.[Abstract/Free Full Text]

53. Behague I, Poirier O, Nicaud V, Evans A, Arveiler D, Luc G, Cambou J-P, Scarabin P-Y, Bara L, Green F, Cambien F. ß fibrinogen gene polymorphisms are associated with plasma fibrinogen and coronary artery disease in patients with myocardial infarction: the ECTIM study. Circulation. 1996;93:440–449.[Abstract/Free Full Text]

54. Montgomery HE, Clarkson P, Nwose OM, Mikailidis DP, Jagroop IA, Dollery C, Moult J, Benhizia F, Deanfield J, Jubb M, World M, McEwan JR, Winder A, Humphries S. The acute rise in plasma fibrinogen concentration with exercise is influenced by the G-453-A polymorphism of the ß-fibrinogen gene. Arterioscler Thromb Vasc Biol. 1996;16:386–391.[Abstract/Free Full Text]

55. Tiret L, Abel L, Rakotovao R. Effect of ignoring genotype-environment interaction on segregation analysis of quantitative traits. Genet Epidemiol. 1993;10:581–586.[Medline] [Order article via Infotrieve]

56. Humphries SE, Cook M, Dubowitz M, Stirling Y, Meade TW. Role of genetic variation at the fibrinogen locus in determination of plasma fibrinogen concentrations. Lancet. 1987;I:1452–1455.

57. Thomas AE, Green FR, Kelleher CH, Wilkes HC, Brennan PJ, Meade TW, Humphries SE. Variation in the promoter region of the ß-fibrinogen gene is associated with plasma fibrinogen levels in smokers and non-smokers. Thromb Haemost. 1991;65:487–490.[Medline] [Order article via Infotrieve]

58. Connor JM, Fowkes FGR, Wood J, Smith FB, Donnan PT, Lowe GDO. Genetic variation at fibrinogen loci and plasma fibrinogen levels. J Med Genet. 1992;27:480–482.[Abstract/Free Full Text]

59. Scarabin P-Y, Bara L, Ricard S, Poirier O, Cambou JP, Arveiler D, Luc G, Evans AE, Samama MM, Cambien F. Genetic variation at the ß-fibrinogen locus in relation to plasma fibrinogen concentrations and risk of myocardial infarction: the ECTIM Study. Arterioscler Thromb. 1993;13:886–891.[Abstract/Free Full Text]

60. Iso H, Folsom AR, Winkelmann JC, Koike K, Harada S, Greenberg B, Sato S, Shimamoto T, Iida M, Komachi Y. Polymorphisms of the beta fibrinogen gene and plasma fibrinogen concentration in Caucasian and Japanese population samples. Thromb Haemost. 1995;73:106–111.[Medline] [Order article via Infotrieve]

61. Heinrich J, Funke H, Rust S, Schulte H, Schonfeld R, Kohler E, Assmann G. Impact of polymorphisms in the alpha- and beta-fibrinogen gene on plasma fibrinogen concentrations of coronary heart disease patients. Thromb Res. 1995;77:209–215.[Medline] [Order article via Infotrieve]

62. de Maat MPM, de Knijff P, Green FR, Thomas AE, Jesperson J, Kluft C. Gender-related associations between ß-fibrinogen genotype and plasma fibrinogen levels and linkage disequilibrium at the fibrinogen locus in Greenland Inuit. Arterioscler Thromb Vasc Biol. 1995;15:856–860.[Abstract/Free Full Text]

63. Thomas AE, Green FR, Lamlum H, Humphries SE. The association of combined and ß fibrinogen genotype on plasma fibrinogen levels in smokers and non-smokers. J Med Genet. 1995;32:585–589.[Abstract/Free Full Text]

64. Thomas AE, Green FR, Humphries SE. Association of genetic variation at the beta-fibrinogen gene locus and plasma fibrinogen levels: interaction between allele frequency of the G/A-455 polymorphism, age and smoking. Clin Genet. 1996;50:184–190.[Medline] [Order article via Infotrieve]

65. Carter AM, Catto AJ, Bamford JM, Grant PJ. Gender-specific associations of the fibrinogen B beta 448 polymorphism, fibrinogen levels, and acute cerebrovascular disease. Arterioscler Thromb Vasc Biol. 1997;17:589–594.[Abstract/Free Full Text]

66. Tracy RP, Bovill EG, Yanez D, Psaty BM, Fried LP, Heiss G, Lee M, Polak JF, Savage PJ. Fibrinogen and factor VIII, but not factor VII, are associated with measures of subclinical cardiovascular disease in the elderly: results from the Cardiovascular Health Study. Arterioscler Thromb Vasc Biol. 1995;15:1269–1279.[Abstract/Free Full Text]

67. Chambless LE, McMahon R, Wu K, Folsom A, Finch A, Shen Y-L. Short-term intraindividual variability in hemostasis factors: the ARIC study. Ann Epidemiol. 1992;2:723–733.[Medline] [Order article via Infotrieve]

68. Nguyen ND, Ghaddar H, Stinson V, Chambless LE, Wu KK. ARIC hemostasis study, IV: intraindividual variability and reliability of hemostatic factors: the Atherosclerosis Risk in Communities (ARIC). Thromb Haemost. 1995;73:256–260.[Medline] [Order article via Infotrieve]

This article has been cited by other articles:

				S. Kathiresan, S. B. Gabriel, Q. Yang, A. L. Lochner, M. G. Larson, D. Levy, G. H. Tofler, J. N. Hirschhorn, and C. J. O'Donnell Comprehensive Survey of Common Genetic Variation at the Plasminogen Activator Inhibitor-1 Locus and Relations to Circulating Plasminogen Activator Inhibitor-1 Levels Circulation, September 20, 2005; 112(12): 1728 - 1735. [Abstract] [Full Text] [PDF]

				M. S. Freeman, M. W. Mansfield, J. H. Barrett, and P. J. Grant Genetic Contribution to Circulating Levels of Hemostatic Factors in Healthy Families With Effects of Known Genetic Polymorphisms on Heritability Arterioscler Thromb Vasc Biol, March 1, 2002; 22(3): 506 - 510. [Abstract] [Full Text] [PDF]

				J. C. Souto, L. Almasy, M. Borrell, M. Gari, E. Martinez, J. Mateo, W. H. Stone, J. Blangero, and J. Fontcuberta Genetic Determinants of Hemostasis Phenotypes in Spanish Families Circulation, April 4, 2000; 101(13): 1546 - 1551. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pankow, J. S. Articles by Sellers, T. A. Search for Related Content PubMed PubMed Citation Articles by Pankow, J. S. Articles by Sellers, T. A.