© 1995 American Heart Association, Inc.
Articles |
Distribution and Correlates of Plasma Fibrinogen in Middle-aged Women
Initial Findings of the Postmenopausal Estrogen/Progestin Interventions (PEPI) Study
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Abstract Fibrinogen levels have been reported in cohort and case-control studies to be positively related to the development of coronary heart disease. This report presents the distribution and determinants of fibrinogen in women enrolling in a 3-year randomized trial of hormone replacement therapy (HRT), the Postmenopausal Estrogen/Progestin Interventions (PEPI) trial. Fasting plasma fibrinogen levels were measured in 874 postmenopausal women, aged 45 to 65 years, who had not used HRT for at least 3.5 months. Mean (±SD) fibrinogen level was 2.83±0.45 g/L. There was a significant positive association between fibrinogen and age (P=.03). Significantly higher (P<.005) fibrinogen levels were seen in current smokers versus nonsmokers (2.94 versus 2.81 g/L), in women who reported consuming fewer than 12 alcoholic drinks in the 12 months before the baseline visit versus higher consumption (2.90 versus 2.79 g/L), and in women who reported never versus ever having used HRT (2.90 versus 2.77 g/L). Self-reported leisure time physical activity (LTPA) was negatively associated (P=.0001) with fibrinogen levels as follows: inactive (2.84 g/L), light (2.89 g/L), moderate (2.80 g/L), and heavy (2.60 g/L), with significantly (P=.0001) lower levels in women who reported heavy LTPA versus each of the other categories and in women reporting moderate versus light LTPA. A strong positive correlation was found between fibrinogen and body mass index (BMI) (r=.32; P<.0001). In a model that included age, smoking, alcohol intake, prior HRT, LTPA, and BMI, LTPA was no longer a statistically significant predictor of fibrinogen level, while associations with other variables remained significant. Fibrinogen was positively associated (P<.001) with waist, hip, and thigh girths and waist-to-hip ratio, but these relationships were no longer significant after adjustment for BMI. In contrast, negative associations with HDL cholesterol (r=-.25; P<.001) and positive associations with LDL cholesterol (r=.16; P<.001) remained significant (P<.01 and P<.05, respectively) after simultaneous adjustment for age, smoking, alcohol, prior HRT, LTPA, and BMI. Statistically significant univariate associations between fibrinogen and triglycerides, insulin and glucose levels before and after oral glucose, and blood pressure were not seen in multivariate analyses. In summary, higher fibrinogen levels were shown to be significantly and independently related to several major lifestyle and physical characteristics known to be associated with increased risk of cardiovascular disease in women.
Key Words: physical exercise • fibrinogen • body composition • lipoproteins • insulin
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In US women, CVD accounts for 48% of deaths, with CHD, the leading cause of death in US women, responsible for half of these and stroke accounting for 9%.1 Thrombosis and hypercoagulability may be involved in atherosclerotic plaque development; therefore, etiologic factors associated with increased risk of thrombotic events in women are of major public health interest. Fibrinogen is perhaps the best studied of the coagulation proteins in clinical conditions and has been shown to be an independent predictor of CVD in several prospective studies of (predominantly white) middle-aged men2 3 4 5 6 7 8 and in the only published prospective study of women.7
In the Framingham Study,7 women in the top tertile (>3.11 g/L) of fibrinogen had twice the risk of CHD over 12 years of follow-up compared with women in the middle and lowest (<2.65 g/L) tertiles. These results are supported by reports from two large population studies; the SHHS9 presented case-control analyses that showed that middle-aged women with a history of myocardial infarction or electrocardiographic changes suggestive of heart disease (n=99) were more likely than matched control subjects to be in the upper quartile of fibrinogen levels (relative risk, 1.35), and the ARIC study10 reported significantly higher fibrinogen levels in both black and white middle-aged women, as well as men, who had prevalent CVD and ultrasound-detected carotid artery intimal-medial thickening compared with participants without CVD. Whether fibrinogen, an acute-phase protein, is a causal risk factor or a marker of CVD is unclear11 ; therefore, it is worthwhile to identify determinants of fibrinogen levels and associations between fibrinogen and other CVD risk factors, which have not been well studied in women.12
Menopause, for instance, may be associated with increases in fibrinogen since higher levels have been found in postmenopausal versus age-matched premenopausal women,13 14 15 and in women undergoing menopause versus women who continue to menstruate,15 16 possibly due to hormonal effects on fibrinogen concentrations. For this reason, the influence of postmenopausal HRT on fibrinogen levels is of particular interest. There is some concern, based on data in women using OCs, which generally contain relatively high-dose synthetic estrogens combined with progestins, that hormonal therapy may increase the risk of both arterial and venous thromboembolic disease, depending on the dose and type of progestin,17 18 and thus may adversely affect the risk of CVD in women who use hormones; however, perimenopausal or postmenopausal women using HRT have been reported to have lower fibrinogen levels than nonusers.15 19 Thus, it is unclear how unopposed estrogens used in low (noncontraceptive) doses affect fibrinogen levels or how adding progestins to a low-dose estrogen regimen influences this hemostatic factor. For these reasons fibrinogen was selected as one of the four primary outcome measures of the PEPI trial, a randomized, double-blinded, placebo-controlled trial of women, aged 45 to 64 years at entry, who were assigned to one of four estrogen-progestin treatments or placebo.20
The purpose of this report is to present a descriptive analysis of fibrinogen levels measured at baseline in the 875 women who were enrolled in PEPI at seven clinics across the United States. Potential correlates of fibrinogen examined here include both behavioral and biological CVD risk factors previously shown to be related to fibrinogen levels. This analysis provides information on relationships between fibrinogen and risk factors in women who are willing to commit to long-term use of estrogen replacement therapy, which has been associated with a decreased risk of CVD.21 22 Since it has been suggested that women who use HRT have differed from women in the general population with respect to socioeconomic and educational level, general health, and lifestyle characteristics, it will be valuable to determine whether relationships between fibrinogen and CVD risk are similar in the PEPI sample to those reported in other population samples. This analysis may also be useful in the interpretation of the recently reported PEPI fibrinogen outcome data.23
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Study Population
Postmenopausal women, aged 45 to 64 years, who had not used HRT for at least 2 months before the first screening visit were recruited for the PEPI trial, as described in detail elsewhere.20 Of 1363 women who completed the first of three screening visits, 875 were randomized into the trial a minimum of 2 months after the first screening visit, thereby extending the minimum time off HRT to 3.5 months before fibrinogen measures were made. Characteristics of the PEPI cohort have been presented elsewhere, along with baseline physical and laboratory methods and values for major end points.20 Important prerandomization exclusions for the purpose of this report include the following: BMI
40
kg/m2; blood pressure (managed by no more than two
antihypertensive agents) >200 mm Hg systolic or >105 mm Hg
diastolic on any one visit or a mean value 160 mm Hg
systolic or 95 mm Hg diastolic on two visits;
fasting glucose level 140 mg/dL (7.77 mmol/L); fasting LDL
cholesterol level 210 mg/dL (5.43 mmol/L) on one visit or
a mean value 190 mg/dL (4.91 mmol/L) on two visits; fasting
triglyceride level 500 mg/dL (5.64 mmol/L) on one visit
or a mean value 400 mg/dL (4.51 mmol/L) on two visits; use of certain
medications, including sodium warfarin or heparin, prescribed
lipid-altering drugs, or high doses of over-the-counter
agents such as fish oils; diabetes requiring insulin therapy; and
history of stroke or transient ischemic attack, thromboembolism
associated with hormone use, heart attack within the previous 6 months,
cancer within the previous 5 years, any breast or endometrial cancer,
or any life-threatening disease.
Physical and Laboratory Methods
Methods for measuring physical parameters (height;
weight; waist, hip, and thigh girths; and resting blood pressure);
collecting fasting blood and urine samples; conducting glucose
tolerance tests; and processing, storing, shipping, and analyzing
biological samples at selected Central Laboratories have been described
in detail elsewhere.20 Briefly, BMI was determined as
weight in kilograms divided by height in meters squared, and BMI
categories were defined according to NHANES-II24 as
"normal weight" (<27.3 kg/m2), "overweight"
(27.3 to 32.2 kg/m2), and "severely overweight"
(>32.2 kg/m2). Single girth measurements included waist
(at the smallest horizontal narrowing between the ribs and iliac crest,
viewed from the front), hip (at the largest horizontal circumference of
the buttocks, viewed from the side), and right thigh (at the largest
circumference with the top of the thigh parallel to the floor, viewed
from the side). WHR categories were defined as "desirable"
(<0.80) and "undesirable" (0.80), according to the recently
revised NCEP ATP-II guidelines.25 Blood pressure was
measured following a modified protocol of the HDFP,26 with
the use of a random-zero sphygmomanometer, after women were seated
undisturbed for at least 5 minutes.
To obtain serum for insulin, (sodium) citrated plasma for fibrinogen, and (sodium) EDTA plasma for glucose and triglyceride and lipoprotein determinations (measured at two separate visits), fasting (12 to 16 hours) blood was collected, after a woman was seated for at least 10 minutes, by nontraumatic venipuncture through a butterfly evacuated tube collection system. Plasma aliquots for fibrinogen, prepared from citrated whole blood samples, were quick-frozen on dry ice. All serum and plasma aliquots were stored at -80°C until they were shipped on dry ice to Central Laboratories. Women were instructed not to do moderate or heavy exercise 12 hours before testing, and if exercise was performed less than 8 hours before testing, the woman was rescheduled. To assess possible acute effects of exercise or use of aspirin and other agents that might affect blood clotting, women were asked about specific activities performed for more than 15 minutes at a time in the 48-hour period preceding a blood draw and about use of aspirin or nonsteroidal anti-inflammatory agents during the 2 weeks before blood collections. Serum and EDTA plasma were also collected at 1 and 2 hours after women ingested 75 g glucose for insulin and glucose, respectively.
Plasma fibrinogen, reported as grams per liter, was measured by a semiautomated version of the Clauss assay27 at the Central Hemostasis Laboratory at the University of Vermont, Burlington. Fibrinogen was measured by rate of clot formation with the use of a BBL Fibrometer (Becton-Dickinson) with the Data-Fi fibrinogen calibration reference plasma (Baxter Healthcare Corp) as the standard. Results were confirmed by participation in the comprehensive coagulation quality assurance program of the College of American Pathologists and by assay of fibrinogen reference material of the College of American Pathologists. The average monthly coefficient of variation for the laboratory's Ci-Trol fibrinogen control plasma (Baxter) was 3.14% for these analyses.
Triglyceride and cholesterol levels were
determined by enzymatic methods28 29 at the Northwest
Lipid Research Laboratories, Seattle, Wash, as were HDL
cholesterol and its subfractions, after precipitation by
dextran sulfate and magnesium.30 31
Cholesterol data are presented according to NCEP
ATP-II cutoff points,25 which specify "desirable,"
"borderline," and "high-risk" levels for total
cholesterol (<200 mg/dL or 5.17 mmol/L; 200 to 239 mg/dL
or 5.17 to 6.21 mmol/L; 240 mg/dL or 6.22 mmol/L) and LDL
cholesterol (<130 mg/dL or 3.36 mmol/L; 130 to 159 mg/dL
or 3.36 to 4.14 mmol/L; and 160 mg/dL or 4.14 mmol/L, with a further
high-risk cutoff point of 190 mg/dL or 4.91 mmol/L, for considering
drug therapy), respectively; and "high-risk" and
"ideal" HDL cholesterol (<35 mg/dL and 60 mg/dL,
respectively). Corresponding NCEP ATP-II cutoff points for
"normal," "borderline," and "high-risk"
triglyceride levels are <200 mg/dL or 2.26 mmol/L; 200 to
399 mg/dL or 2.26 to 4.52 mmol/L; and 400 mg/dL or 4.52 mmol/L,
respectively, with greater concern for levels 1000 mg/dL or 11.3
mmol/L. Insulin was determined by immunoassay with the use of a
two-antibody system32 at Indiana University School of
Medicine, Indianapolis, where glucose was also
assayed.33
Questionnaire Information
Extensive information was gathered by questionnaire for each
PEPI participant. Data ascertained included health-related
behaviors (smoking, alcohol consumption, physical activity level,
diet), psychosocial factors (depressive symptomatology, sexual
functioning), medical and reproductive history, including
history of OC and HRT use, current and past medication use, family
history of CVD, and demographic information (marital status,
educational level, occupation). Smoking questions identified current
smokers and women with a history of smoking, including the age of onset
and cessation, if applicable, and how many cigarettes a woman smoked
per day. The number of pack-years was estimated by multiplying
current age minus age at start of smoking by the number of cigarettes
smoked daily, divided by 20, ie, the number of cigarettes per pack.
Questions regarding alcohol intake were designed to separate never or
extremely light drinking from more regular intake, including number of
days alcohol was consumed each month and the average amount consumed on
such a day.
Exercise questions included measures of self-reported physical activity levels at work, in the home, and at leisure time during the 12 months before the baseline visit and required the woman to rate her activity level as inactive, light, moderate, or heavy, with the use of generally accepted descriptions for these terms: for inactive, the person spends most waking hours sitting or standing quietly; for light, the person usually walks for more than 10 minutes at a time each day, leisurely rides a bicycle, fishes, bowls, golfs, and/or engages in light chores; for moderate, the person participates in brisk walking, recreational or doubles tennis, and swimming and engages in moderate chores; and for heavy activity, the person performs vigorous activity on a regular basis, including jogging, singles tennis, high-intensity aerobics, and/or heavy chores.
Statistical Procedures
Data were analyzed with the use of general linear
modeling techniques including correlational analysis,
regression, and ANOVA. ANCOVA determined mean fibrinogen levels
adjusted for important covariates. Calculations were performed on the
SAS system. A multiple regression model was developed to
determine predictors of fibrinogen in the PEPI women in which
variables that were seen to be strongly associated with fibrinogen
were removed by backward stepwise procedures.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
As reported elsewhere,20 the PEPI cohort, which is 89% white, is composed of women of relatively high educational level and socioeconomic status. The cohort had a mean (±SD) age of 56.1±4.3 years at randomization and reported a 69% prevalence of natural menopause, a 32% prevalence of hysterectomy, and a 15% prevalence of oophorectomy (unilateral or bilateral). Ninety percent had at least one pregnancy, 60% reported prior OC use, and 53% had used HRT, with concomitant progestin therapy in 63% of estrogen users. Approximately half (50.6%) of PEPI women reported ever having smoked cigarettes regularly, but only 13% currently smoked. More than two thirds (68.2%) had consumed at least 12 alcoholic beverages in the last 12 months. More than one half reported at least a moderate level of physical activity at home (50.5%) and in their leisure time (51.2%), and nearly one third (30.4%) reported moderate or heavy physical activity at work. All but 2% rated their general health as good (26%), very good (44%), or excellent (29%).
The laboratory fibrinogen value for one participant was considered physiologically implausible (0.05 g/L), and her data were excluded, leaving 874 women for these analyses. The mean fibrinogen level was 2.83±0.45 g/L (±SD), with the 10th, 25th, 50th, 75th, and 90th percentile values being 2.27, 2.52, 2.79, 3.14, and 3.40 g/L, respectively. Although fibrinogen levels were approximately 5% higher in nonwhite than white women, no significant differences were seen between mean (±SD) fibrinogen levels in whites (2.81±0.45; n=774) and any ethnic group: Hispanic (2.94±0.46; n=47), black (2.93±0.58; n=32), Asian (2.99±0.34; n=17), or Native American (2.97±0.63; n=4).
Table 1
presents associations between mean
fibrinogen levels and age, BMI, and selected behavioral factors:
smoking, alcohol intake, physical activity level, and history of HRT
use. Fibrinogen levels increased significantly with age across the
20-year age span of the PEPI cohort and correlated significantly
(P=.001) and positively with BMI (r=.32). Current
smokers had higher mean fibrinogen levels than nonsmokers; however,
fibrinogen values did not differ (P=.86) between women who
ever smoked regularly in their lifetime (2.82 g/L; n=443), which
included current smokers, and women who never smoked (2.83 g/L; n=431).
In current smokers, neither the number of cigarettes smoked per day,
which averaged 16.9±10.5 (SD), nor pack-years of exposure, which
averaged 30, was related to fibrinogen levels (P=.29 and
P=.18, respectively). No differences were seen between women
who reported consuming fewer than 12 drinks ever in their lifetime or
fewer than 12 drinks in any year versus higher consumption; however,
women who reported consuming fewer than 12 alcoholic beverages in the
12 months before the baseline visit had significantly higher fibrinogen
levels than those reporting higher alcohol intake in that year. Among
subjects who consumed more than 12 drinks during this period, drinking
was reported for an average of 11 days per month, with a mean of 1.7
drinks consumed per drinking day, or an average of 20 drinks per month.
While the number of drinks per drinking day was not related to
fibrinogen level in these women, both the number of drinking days per
month and the total number of drinks per month were significantly and
inversely related to fibrinogen level (P=.0001).
|
The level of physical activity reported at work (P=.30) was
not associated with fibrinogen level; however, higher activity levels
at home were associated with lower fibrinogen levels
(P=.04), and the level of physical activity reported in
leisure time was inversely related to mean fibrinogen values, as shown
in Table 1
. Women who reported heavy leisure time activity had
significantly lower levels than inactive women (P=.007),
women reporting light activity (P=.0001), and women
reporting moderate activity (P=.004); and, moderately active
women had lower levels than those reporting light activity in leisure
time (P=.004). Simultaneously adjusting the
relationship between fibrinogen and physical activity for age, smoking,
alcohol intake, and prior HRT had little effect on this association
(P=.0005); however, addition of BMI to the analysis
eliminated the significance of this relationship
(P=.11).
An exercise session of at least 15 minutes' duration during the 12 to 48 hours before testing was reported by 332 women, who had lower fibrinogen levels than the 541 women who had not exercised in the 2 days before testing (2.78 versus 2.85 g/L; P=.02). Women who reported heavy exercise during this period had lower levels (2.68 g/L; n=83) than women who reported light (2.81 g/L; n=66; P=.07) or moderate (2.82 g/L; n=183; P=.03) exercise. In women reporting recent exercise, fibrinogen levels were not related to the length of the exercise session, which averaged 74 minutes and ranged from 10 to 720 minutes (P=.28), or to the length of time since the exercise was performed relative to testing (P=.48).
As shown in Table 1, higher fibrinogen levels were found in women who
reported never having used versus ever having used HRT, but neither
length of time on HRT (P=.49) nor progestin therapy
(P=.18), reported by 60% of estrogen users, significantly
influenced fibrinogen levels.
After simultaneous adjustment of relationships between
fibrinogen and the selected CVD risk factors presented in Table 1 for age, BMI, current smoking (yes or no), intake of 12 or more
alcoholic beverages in the past 12 months (yes or no), reported leisure
time activity level (by four levels), and prior HRT use (yes or no),
fibrinogen remained significantly associated with age, BMI, smoking,
alcohol intake, and prior HRT use but was no longer associated with
physical activity.
Fig 1 presents fibrinogen levels by the BMI
categories defined in NHANES-II, with a further breakdown of the
nonobese category for BMI <19 kg/m2, which is often
regarded as "clinically underweight" and by the WHR categories
defined in NCEP ATP-II. The significant difference in fibrinogen levels
between WHR groups (P=.0001) was eliminated by adjusting for
BMI (P=.18), whereas adjusting the relationship between
fibrinogen and BMI for WHR had no effect on the significance of that
relationship. Fibrinogen levels also increased significantly
(P<.001) across tertiles of waist, hip, and thigh girths,
as shown in Fig 2; however, levels did not differ across
girth tertiles after adjusting for BMI (P=.06,
P=.63, and P=.81, respectively).
|
|
Table 2 presents mean fibrinogen levels by selected
medical and family history parameters. Fibrinogen levels
did not differ between women who reported a history of OC use versus
never using OCs or among subjects with surgical menopause, natural
menopause less than 5 years, or natural menopause more than 5 years
before testing. Fibrinogen levels also did not differ between women who
did or did not use aspirin or nonsteroidal anti-inflammatory agents
during the 2 weeks before blood collection. Fibrinogen levels were not
related to either hematocrit (r=.01) or hemoglobin
(r=-.01) in the PEPI cohort, which had a mean
hematocrit of 40.6±2.7% (SD) and mean hemoglobin of 13.6±0.9 g/dL.
Fibrinogen levels did not differ between subjects who reported a
history of hypertension in either a parent or sibling versus no such
family history, with women for whom this was unknown (n=142) excluded
from the analysis, nor did levels differ between women
who reported a history of myocardial infarction in a parent or a
sibling versus no such family history, with women for whom this was
unknown (n=42) excluded from the analysis.
|
Table 3 presents the mean baseline values for
triglyceride and lipoprotein variables, glucose and
insulin (fasting and after glucose load), systolic and
diastolic blood pressure variables, and correlations
for these variables with initial fibrinogen levels, before and
after simultaneous adjustment for age, smoking, alcohol,
leisure time activity, prior HRT use, and BMI. In
univariate analyses, fibrinogen levels were
significantly related to all of the major CHD risk factors measured;
however, after we adjusted for key behavioral CHD risk factors,
fibrinogen remained significantly and negatively related to HDL
cholesterol, HDL subfractions, and apolipoprotein A-I and
positively associated with LDL cholesterol and
apolipoprotein B but was no longer related to total
cholesterol or triglycerides, glucose or
insulin, or blood pressure, suggesting that relationships between
fibrinogen and these CHD risk factors are not independent of their
associations with other, primarily lifestyle, risk factors.
|
Fig 3 presents fibrinogen levels by NCEP cutoff
points, with HDL cholesterol further separated at 50 mg/dL
or 1.29 mmol/L, a cutoff point that was shown to distinguish risk of
CHD in women in the Lipid Research Clinics Follow-up
Study.34 Fibrinogen levels were significantly higher
(P=.0001) as total and LDL cholesterol levels
rose from the desirable to the high-risk categories and as HDL
cholesterol levels decreased from ideal to high-risk
levels. Fasting triglyceride levels were in the desirable
range in 812 women (93% of the sample) and reached the high-risk
level in only two women; therefore, these data are not
presented. However, as triglyceride levels
increased from <100 mg/dL or 1.13 mmol/L (n=525) to 100 to 199 mg/dL
or 1.13 to 2.26 mmol/L (n=287) and to 200 mg/dL or 2.26 mmol/L
(n=62), significantly higher fibrinogen levels were seen (2.76, 2.92,
and 2.91 g/L, respectively; P=.0001).
|
Fig 4 compares mean fibrinogen levels in women who fall
into specific categories related to carbohydrate
metabolism, including fasting glucose <110 mg/dL versus
110 mg/dL (6.11 mmol/L) and 2-hour postload glucose <200 mg/dL
versus 200 mg/dL (11.10 mmol/L), which are often regarded clinically
as cutoff points between desirable versus elevated blood sugar. Fig 4
also shows fibrinogen levels according to PEPI tertiles for fasting and
2-hour postload insulin levels. Fibrinogen levels were significantly
increased in women with higher fasting postload glucose and insulin
versus lower levels.
|
Fig 5 compares mean fibrinogen levels in women with
systolic blood pressure <140 mm Hg versus 140 mm Hg and
diastolic blood pressure <90 mm Hg versus 90 mm Hg,
current cutoff points used to define hypertension.35
Fibrinogen levels were significantly higher in women with higher
systolic and diastolic blood pressures.
|
Table 4 presents a multiple regression model that
includes the main variables that were significantly associated with
fibrinogen levels in Tables 1 through 3: age, BMI, current smoking
status, alcohol intake (12 drinks during the previous year), history
of HRT use, HDL cholesterol, and LDL
cholesterol. Fibrinogen levels increased 0.01 g/L for each
year of age and 0.03 g/L for each increase in BMI unit
(kg/m2) and were 0.15 g/L higher in current smokers versus
nonsmokers, 0.06 g/L higher in women who had consumed fewer than 12
alcoholic drinks in the prior year versus more, and 0.09 g/L higher in
women with no history of prior postmenopausal HRT. Fibrinogen levels
were also 0.05 g/L higher per millimole per liter of LDL
cholesterol increase and 0.10 g/L lower per millimole per
liter increase in HDL cholesterol in PEPI women.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Fasting plasma fibrinogen levels in 874 middle-aged, relatively healthy women participating in the PEPI trial were significantly higher with increasing age, BMI, and current cigarette smoking and lower with moderate alcohol consumption and history of prior HRT use. These findings generally support previous reports of age and behavioral correlates of fibrinogen in other epidemiological samples of predominantly white women, including the 761 women in the Framingham Study,7 the SHHS sample of approximately 4300 women aged 40 to 59 years,13 the large ARIC cohort of approximately 6500 women aged 45 to 64 years,19 the second MONICA Augsburg survey of 2152 women aged 25 to 74 years,36 and the Muenster Arteriosclerosis Study of approximately 1300 women.37 In each study, significantly higher fibrinogen levels were seen with increasing age, relative body weight, and cigarette smoking, whereas alcohol was generally associated with lower levels, except in the Muenster Arteriosclerosis Study cohort.34 In ARIC women, regression models predicted a 0.008-g/L increase in fibrinogen for each year of age, an approximately 0.02-g/L increase for each BMI unit, and a 0.13-g/L higher level in smokers.19 These predictions are reasonably close to our findings of a 0.01-g/L fibrinogen increase per year of age, a 0.03-g/L increase with each BMI unit, and a 0.15-mg/dL higher level in smokers in the PEPI cohort; however, we recognize that these results are specific to this particular sample and may not be generalizable.
ARIC also predicted a 0.03- to 0.04-g/L increase per unit of a sport index that ranged from 1 (low) to 5 (high), corresponding to a 0.15- to 0.20-mg/dL difference between the lowest and highest sport index.19 In PEPI, women reporting inactivity or light activity in leisure time had fibrinogen levels that were approximately 0.25 g/L higher than women reporting heavy leisure time activity, and a dose-response relationship was seen for the association of activity level in leisure time; however, this relationship was not independent of BMI. This finding should not be interpreted as evidence that exercise has no influence on fibrinogen, however, because exercise may directly cause reductions in body weight, whereas inactivity is generally associated with a higher BMI.38 Performance of heavy exercise in the 48-hour period before blood testing was also associated with lower fibrinogen levels; however, it is probable that women who routinely engage in heavy activity will do so within the 2-day period before testing, and therefore differences in fibrinogen levels between women who did versus did not report recent exercise probably reflects the influence of habitual activity, which includes lower body weight, rather than an acute effect of exercise on fibrinogen.
As recently reviewed by Ernst,39 several longitudinal studies in men suggest that endurance exercise for several weeks will decrease fibrinogen levels by 0.4 g/L. The mechanism by which exercise could alter fibrinogen has not been well investigated. Production of fibrinogen in the liver may be altered during exercise by exposure of hepatic cells to metabolites associated with exercise, including free fatty acids, by changes in hormonal balance, and/or by decreased splanchnic blood flow. Exercise-induced changes in blood pressure or vascular physiology may also alter hemostatic dynamics that influence fibrinogen metabolism. It is possible that the primary means by which routine exercise levels influence fibrinogen is by reducing body fat; however, the means by which excess body fat might affect fibrinogen is also unclear, although higher body fat is associated with specific changes in activities of key enzymes in hepatocytes, adipocytes, and skeletal muscle fibers.
It is well known that obesity is related to other CHD risk factors in women, including low HDL cholesterol, high triglyceride levels, impaired glucose tolerance, and elevated blood pressure,40 but the physiological mechanisms underlying these relationships are not well understood. In PEPI, fibrinogen was shown to be higher in women with lower HDL cholesterol and higher fasting and 2-hour glucose and insulin levels and in hypertensive versus normotensive women; however, only the relationship with HDL cholesterol remained significant after we adjusted for age and BMI. The Framingham Study7 reported significantly higher fibrinogen levels with increasing total serum cholesterol level, hypertension, and glucose intolerance; however, none of these findings were shown to be independent of other factors in PEPI women.
Several investigators have reported a link between body build and fibrinolytic variables, but WHR was not associated with fibrinogen levels in these studies.12 On the other hand, MONICA investigators reported an independent positive association between fibrinogen and WHR in women as well as in men.36 In PEPI, strong significant associations between fibrinogen and WHR, waist-to-thigh girth ratio, and waist, hip, and thigh girths were eliminated with adjustment for BMI. Although it has been questioned whether obesity level (generally reflected by BMI) is as important in determining the metabolic consequences of obesity as the deposition of excess body fat in intra-abdominal rather than subcutaneous adipose tissue, there is likely an important interaction between the amount of excess fat and its distribution. In a small study of obese, middle-aged women recruited from an obesity outpatient clinic, fibrinogen levels were elevated in obese women with a high WHR compared with those with a low WHR and also compared with lean women with a low or high WHR.41 Therefore, although having excess centrally located fat may be even more detrimental than having excess fat in the more common "female" sites (subcutaneous thigh and upper arm depots), presumably due to the release of fatty acids from intra-abdominal fat depots into the portal vein, where they reach the liver in greater concentrations than fatty acids released from subcutaneous fat depots, the amount of excess fat stored is clearly of great importance.
Fibrinogen was higher in PEPI women who had previously used HRT, as also reported by the ARIC investigators.19 However, history of OC use was not related to current fibrinogen levels in PEPI, in contrast to the finding in the SHHS that women with a history of OC use had lower levels than women who had not used OCs.13 This latter difference may arise because all PEPI women were postmenopausal, whereas a substantial number of SHHS women were premenopausal. The association of fibrinogen with previous HRT raises the obvious question of possible lingering effects of HRT versus other differences that may influence fibrinogen levels between women who use HRT compared with nonusers, including behavioral variables such as smoking and activity level.
ARIC women showed increases in fibrinogen across quartiles of triglycerides, LDL cholesterol, and fasting insulin and decreases in fibrinogen across HDL cholesterol quartiles.19 PEPI women also showed higher fibrinogen levels as triglyceride, total and LDL cholesterol, and fasting and 2-hour glucose and insulin levels increased across clinically relevant categories, whereas lower fibrinogen levels were found as HDL cholesterol increased. The relationship between fibrinogen and fasting triglyceride levels was eliminated after we adjusted for confounding variables; however, 93% of PEPI women had triglycerides <200 mg/dL, and it is clearly possible that strong relationships between fibrinogen and triglycerides can only be seen at higher triglyceride values or with greater variation in levels within the study sample. The relationship between fibrinogen and HDL and LDL cholesterol remained significant after simultaneous adjustment for age, BMI, smoking, alcohol intake, leisure time exercise level, and history of HRT. Possible mechanisms that may underlie these independent associations have not been elucidated. LDL cholesterol is generally believed to be influenced predominantly by diet composition, particularly saturated fat intake, while HDL cholesterol is generally believed to be affected by obesity status, activity level, smoking, and fat intake.
In summary, higher fibrinogen levels were shown to be independently related to several lifestyle and physical characteristics known to be associated with risk of CVD in women, including age, obesity status, smoking, alcohol intake, history of HRT, HDL and LDL cholesterol, and their associated apolipoproteins A-I and B, respectively. In contrast, the strong relationships between fibrinogen concentrations and physical activity level and girths (including WHR) were not independent of BMI. The independent association of fibrinogen with so many CHD risk factors suggests that although fibrinogen may be a predictor of CHD, it may not be a biochemically or physiologically independent risk factor for CHD. It would be valuable to determine whether changing body fat or key lifestyle factors, such as smoking, alcohol intake, or activity level, alter fibrinogen levels. It would also be worthwhile to determine which behavioral variables and physiological factors may underlie associations between fibrinogen and the lipoprotein profile. Finally, it would be valuable to learn whether lowering fibrinogen levels reduces a woman's risk for CHD.
|
Selected Abbreviations and Acronyms |
|---|
|
|
Acknowledgments |
|---|
The PEPI trial was supported by cooperative agreement research grants (U01-HL40154, U01-HL40185, U01-HL40195, U01-HL40205, U01-HL40207, U01-HL40231, U01-HL40232, and U01-HL40273) from the National Heart, Lung, and Blood Institute; the National Institute of Child Health and Human Development; the National Institute of Arthritis and Musculoskeletal and Skin Diseases; the National Institute of Diabetes and Digestive and Kidney Diseases; and the National Institute on Aging.
|
Footnotes |
|---|
Reprint requests to Marcia L. Stefanick, PhD, Stanford Center for Research in Disease Prevention, 730 Welch Rd, Suite B, Palo Alto, CA 94304-1583.
From Stanford University (Calif) (M.L.S.); Bowman Gray School of Medicine, Wake Forest University, Winston-Salem, NC (C.L., G.H.); University of Vermont, Burlington (R.P.T.); George Washington University, Washington, DC (C.M.K.); Robert Wood Johnson Medical School, Piscataway, NJ (D.L.L.); and Johns Hopkins University, Baltimore, Md (T.L.B.).
Received May 12, 1995; accepted September 27, 1995.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Wenger NK, Speroff L, Packard B, eds. Proceedings of an NHLBI Conference: Cardiovascular Health and Disease in Women. Greenwich, Ct: Le Jacq Communications, Inc; 1993.
2. Meade TW, Brozovic M, Chakrabarti RR, Haines AP, Imeson JD, Mellows S, Miller GJ, North WRS, Stirling Y, Thompson SG. Haemostatic function and ischaemic heart disease: principal results of the Northwick Park Heart Study. Lancet. 1986;1:533-537. [Medline] [Order article via Infotrieve]
3.
Yarnell JWG, Baker IA, Sweetnam PM, Bainton D,
O'Brien JR, Whitehead PJ, Elwood PC. Fibrinogen, viscosity, and
white blood cell count are major risk factors for ischemic
heart disease: the Caerphilly and Speedwell Collaborative Heart Disease
Studies. Circulation. 1991;83:836-844.
4. Wilhelmson L, Svardsudd K, Korsan-Bengtsen K, Larsson B, Welin L, Tibblin G. Fibrinogen as a risk factor for stroke and myocardial infarction. N Engl J Med. 1984;311:501-505. [Abstract]
5. Eriksson H, Korsan-Bengtsen K, Svardsudd WL, Larsson B, Tibblin G. Wilhelmsen L. 21-year follow up of CVD and total mortality among men born in 1913. In: Ernst E, Koenig W, Lowe GDO, Meade TW, eds. Fibrinogen: A `New' Cardiovascular Risk Factor. Vienna, Austria: Blackwell; 1992:115-119.
6. Stone MC, Thorp JM. Plasma fibrinogen: a major coronary risk factor. J R Coll Gen Pract. 1985;35:565-569. [Medline] [Order article via Infotrieve]
7.
Kannel WB, Wolf PA, Castelli WP, D'Agnostino
RB. Fibrinogen and risk of cardiovascular
disease: the Framingham Study. JAMA. 1987;258:1183-1186.
8.
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.
9. Lee AJ, Lowe GDO, Woodward M, Tunstall-Pedoe H. Plasma fibrinogen and indicators of coronary heart disease. Perfusion. 1991;4:432. Abstract.
10.
Folsom AR, Wu KK, Shahar E, Davis CE, for the
Atherosclerosis Risk in Communities (ARIC) Study
Investigators. Association of hemostatic variables with prevalent
cardiovascular disease and asymptomatic
carotid artery atherosclerosis.
Arterioscler Thromb. 1993;13:1829-1836.
11.
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.
12. Meilahn EN. Hemostatic factors and risk of cardiovascular disease in women: an overview. Arch Pathol Lab Med. 1992;116:1313-1317. [Medline] [Order article via Infotrieve]
13. Lee AJ, Smith WCS, Lowe GDO, Tunstall-Pedoe H. Plasma fibrinogen and coronary risk factors: the Scottish Heart Health Study. J Clin Epidemiol. 1990;43:913-919. [Medline] [Order article via Infotrieve]
14. Meade TW, Imeson JD, Haines AP, Stirling Y, Thompson SG. Menopausal status and haemostatic variables. Lancet. 1983;1:22-24. [Medline] [Order article via Infotrieve]
15. Meilahn EN, Kuller LH, Matthews KA, Kiss JE. Hemostatic factors according to menopausal status and use of hormone replacement therapy. Ann Epidemiol. 1992;2:445-455. [Medline] [Order article via Infotrieve]
16. Meade TW, Dyer S, Howarth DJ, Imeson JD, Stirling Y. Antithrombin III and procoagulant activity: sex differences and effects of menopause. Br J Haematol. 1990;74:77-81. [Medline] [Order article via Infotrieve]
17. Meade TW. Risks and mechanisms of cardiovascular events in users of oral contraceptives. Am J Obstet Gynecol. 1988;158:1646-1652. [Medline] [Order article via Infotrieve]
18. Task Force on Oral Contraceptives-WHO Special Programme of Research, Development and Research Training in Human Reproduction. A multicentre study of coagulation and haemostatic variables during oral contraception: variations with four formulations. Br J Obstet Gynecol. 1991;98:1117-1128. [Medline] [Order article via Infotrieve]
19. Folsom AR, Conlan MG, Davis CE, Wu KK, for the Atherosclerosis Risk in Communities (ARIC) Study Investigators. Relations between hemostatic variables and cardiovascular risk factors in middle-aged adults. Ann Epidemiol. 1992;2:481-494. [Medline] [Order article via Infotrieve]
20. The Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial Investigators. The Postmenopausal Estrogen/Progestin Interventions (PEPI) trial. Control Clin Trials. 1995;16 (No. 4S).
21.
Bush TL, Barrett-Connor E, Cowan LD, Cowan LD, Criqui
MH, Wallace RB, Suchindran CM, Tyroler HA, Rifkind BM.
Cardiovascular mortality and noncontraceptive use of
estrogen in women: results from the Lipid Research Clinics Program
Follow-up Study. Circulation. 1987;75:1102-1109.
22. Stampfer MJ, Willett WC, Colditz GA, Rosner B, Speizer FE, Hennekens CH. A prospective study of postmenopausal estrogen therapy and coronary heart disease. N Engl J Med. 1985;313:1044-1049. [Abstract]
23.
The Writing Group for the PEPI Trial. Effects of
estrogen or estrogen/progestin regimens on heart disease risk factors
in postmenopausal women: the Postmenopausal Estrogen/Progestin
Interventions (PEPI) trial. JAMA. 1995;273:199-208.
24. National Center for Health Statistics. Plan and Operation of the Second National Health and Nutrition Examination Survey, 1976-1980: Vital and Health Statistics. Washington, DC: US Government Printing Office; 1981. US Dept of Health, Education, and Welfare publication PHS 81-1317, series 1, No. 15.
25.
Expert Panel on Detection, Evaluation, and Treatment of
High Blood Cholesterol in Adults. Summary of the second
report of the National Cholesterol Education Program (NCEP)
Expert Panel on Detection, Evaluation, and Treatment of High Blood
Cholesterol in Adults (Adult Treatment Panel II).
JAMA. 1993;269:3015-3023.
26. Labarthe DR, Poizner SB, Cutter GR, Casey BH. Measurement of blood pressure: a manual for training and certification of observers, based on the procedures of the Hypertension Detection and Follow-up Program. Prev Med. 1976;5:207-315. [Medline] [Order article via Infotrieve]
27. Clauss A. Gerinnungs physiologishce schnellmethode zur bestimmung des fibrinogens. Acta Haematol. 1957;17:237-246. [Medline] [Order article via Infotrieve]
28. Hainline A Jr, Karon J, Lippel K, eds. Manual of Laboratory Operations: Lipid and Lipoprotein Analysis. 2nd ed. Bethesda, Md: US Dept of Health and Human Services, Public Health Service, National Institutes of Health; 1983.
29. Warnick GR. Enzymatic methods for quantification of lipoprotein lipids. In: Albers JJ, Segrest JP, eds. Methods in Enzymology. Orlando, Fla: Academic Press, Inc; 1986;129:101-123.
30. Warnick RG, Benderson JM, Albers JJ. Quantification of high-density-lipoprotein subclasses after separation by dextran sulfate and Mg2+ precipitation. Clin Chem. 1982;28:1574. Abstract.
31. Bachorik PS, Albers JJ. Precipitation methods for quantification of lipids. In: Albers JJ, Segrest JP, eds. Methods in Enzymology. Orlando, Fla: Academic Press, Inc; 1986;129:78-100.
32. Morgan CR, Lazarow WA. Immunoassay of insulin: two antibody system for plasma insulin levels of normal subdiabetic and diabetic rats. Diabetes. 1963;12:115-126.
33.
Somogyi M. Notes on sugar
determinations. J Biol Chem. 1952;195:19-23.
34.
Bass KM, Newschaffer CJ, Klag MJ, Bush TL.
Plasma lipoprotein levels as predictors of
cardiovascular death in women. Arch
Intern Med. 1993;153:2209-2216.
35. National High Blood Pressure Education Program. Working group report on primary prevention of hypertension. Bethesda, Md: National Heart, Lung, and Blood Institute; 1993. US Public Health Service, Dept of Health and Human Services publication NIH 93-2669.
36.
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: results
from the Second MONICA Augsburg Survey, 1989-1990.
Arterioscler Thromb. 1992;12:780-788.
37. 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]
38. Stefanick ML. Exercise and weight control. In: Holloszy JO, ed. Exercise and Sport Science Reviews. Baltimore, Md: Williams & Wilkins; 1993;21:363-396. American College of Sports Medicine Series.
39.
Ernst E. Regular exercise reduces fibrinogen
levels: a review of longitudinal studies. Br J Sports
Med.. 1993;27:175-176.
40. Stefanick ML. The roles of obesity, regional adiposity, and physical activity in coronary heart disease in women. In: Wenger NK, Speroff L, Packard B, eds. Proceedings of an NHLBI Conference: Cardiovascular Health and Disease in Women. Greenwich, Ct: Le Jacq Communications, Inc; 1993:149-156.
41. Landin K, Stigendal L, Eriksson E, Krokeiwski M, Risberg G, Tengborn L, Smith U. Abdominal obesity is associated with an impaired fibrinolytic activity and elevated plasminogen activator inhibitor-1. Metabolism. 1990;39:1044-1048.[Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  L. M. Szymanski, C. M. Kessler, and B. Fernhall Relationship of physical fitness, hormone replacement therapy, and hemostatic risk factors in postmenopausal women J Appl Physiol, April 1, 2005; 98(4): 1341 - 1348. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. D. P. Luyer, S. Khosla, W. G. Owen, and V. M. Miller Prospective Randomized Study of Effects of Unopposed Estrogen Replacement Therapy on Markers of Coagulation and Inflammation in Postmenopausal Women J. Clin. Endocrinol. Metab., August 1, 2001; 86(8): 3629 - 3634. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Eriksson, N. Egberg, S. Wamala, K. Orth-Gomer, M. A. Mittleman, and K. Schenck-Gustafsson Relationship Between Plasma Fibrinogen and Coronary Heart Disease in Women Arterioscler Thromb Vasc Biol, January 1, 1999; 19(1): 67 - 72. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. DeSouza, P. P. Jones, and D. R. Seals Physical Activity Status and Adverse Age-Related Differences in Coagulation and Fibrinolytic Factors in Women Arterioscler Thromb Vasc Biol, March 1, 1998; 18(3): 362 - 368. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||