SEARCH

SEARCH BY CITATION

Keywords:

  • critical limb ischaemia;
  • fibrinogen;
  • peripheral arterial occlusive disease;
  • thrombosis

Abstract.

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Objective. Plasma fibrinogen predicts cardiovascular events in patients with stable peripheral arterial occlusive disease, but its predictive value in patients with chronic critical limb ischaemia, a condition associated with a high risk of death, is unknown.

Design. A prospective cohort study.

Setting. Determination of fibrinogen and other potential predictors during clinic-based work-up of patients admitted for diagnostic and therapeutic evaluation.

Subjects. A total of 108 patients (72 ± 10 years, 78 males) with atherosclerotic occlusive disease and critical limb ischaemia (pain at rest and/or trophic lesions) followed up for a median period of 1.6 years (range: 8 days–5.5 years; 218 patient-years).

Main outcome measures. Total mortality.

Results. Forty-five deaths (71% cardiovascular) occurred during the follow-up. Baseline fibrinogen was higher in those who died in the early follow-up period (first 6 months), as were white cell count and serum creatinine, while haematocrit was lower. Plasma fibrinogen values correlated positively with white cell count, and negatively with haematocrit; other cardiovascular prognostic factors did not differ. Only plasma fibrinogen predicted survival independently in multivariate age-corrected Cox regression analysis. Relative risk of death doubled for each standard deviation above the mean and increased with each tertile increase in fibrinogen.

Conclusions. Fibrinogen predicted death in these elderly arteriopaths with critical limb ischaemia, particularly those who died in the first months following critical ischaemia. Inflammatory stimuli secondary to severely defective tissue oxygenation and possibly sepsis and necrosis, might have stimulated fibrinogen, an acute-phase reactant, thereby compromising organ perfusion through increased blood viscosity and/or promoting thrombosis.


Introduction

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Plasma fibrinogen levels are consistently increased in patients with stable peripheral arterial occlusive disease [ e.g. 1–5] – the higher the concentration, the worse the extent of vascular disease as determined by angiography [ 6]. This finding has prognostic implications, as raised fibrinogen concentrations have been found in still healthy subjects who subsequently progressed to peripheral vascular disease [ 7]. Similarly, prospective observation of patients with clinically evident peripheral arterial occlusive disease demonstrated a close link between elevated baseline fibrinogen levels and progression of peripheral arterial disease [ 8], graft occlusion after femoropopliteal vein bypass [ 9] and mortality, both cardiovascular and overall [ 8, 10, 11]. Thus, by now the role of fibrinogen as a risk factor for an adverse evolution of stable peripheral arterial occlusive disease seems established [ 12]. However, no information exists on fibrinogen as a prognostic predictor in that portion of patients [ 13, 14] in whom arterial disease has progressed to chronic critical limb ischaemia [ 15], a clinical condition characterized by elevated risk of death [ 16–18].

Materials and methods

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Protocol

The cohort included 108 consecutive, mainly elderly (85% > 65 years) subjects recruited between 28 January 1991 and 6 July 1996 from the area of Pisa and neighbouring geographical districts of Leghorn and Lucca. Patients were admitted for diagnostic screening (including angiography in all, as well as echocardiogram and cardiac stress tests in most, followed by coronary angiography as deemed necessary) for atherosclerotic peripheral arterial occlusive disease, chronic critical limb ischaemia and associated diseases. The diagnostic criteria for chronic critical limb ischaemia were defined as pain at rest for at least 15 days, requiring intensive analgesic treatment (= 27, four females, stage Fontaine III, age: 70 ± 9 years) and/or trophic lesions (= 81, 26 females, stage Fontaine IV, age: 72 ± 10 years). After discharge from the clinic, 79 patients underwent major limb salvage surgery (= 62) and/or amputation (= 17), without any intra- or perioperative deaths. Patients in whom surgery was technically impossible (= 29) were referred for antalgic treatment and/or spinal cord stimulation.

The single criterion for inclusion in the study was the presence of angiographically assessed atherosclerotic peripheral arterial occlusive disease; patients on oral anticoagulants or fibrates at baseline were excluded. Follow-up started upon admission to the clinic and survival was checked by telephone yearly from 1991 to 1996. Death from any cause or final ascertainment of survival in the last week of July 1996 ended the census. One patient was lost to follow-up after the first annual screening and censored accordingly. Final diagnosis was assessed through death certificates, verified in most by review of medical records and/or interviews with family members and the general practitioner, and/or autopsy records (= 3). Fatalities were classified as (1) cardiovascular, (2) neoplastic, and (3) miscellaneous. Patients and families were aware of the reason for the inquiry and agreed to participate.

Laboratory methods

Blood samples for plasma fibrinogen concentrations were drawn after an overnight fast in tubes containing trisodium citrate 0.13 mm L−1 (one volume per nine of blood). Fibrinogen was measured photometrically (Electra 1000C, Medical Laboratory Automation Inc., Pleasantville, USA) according to the prothrombin time-derived method (Dade SpA, Roma, Italy) [ 19]. The internal reference plasma samples used to assess reproducibility showed a mean monthly variation coefficient of 4.8%. Total cholesterol and triglycerides were assayed by enzymatic colorimetric techniques (Boehringer-Mannheim, Mannheim, Germany), white cell count (WCC) and haematocrit by automated methods (H1 Technicon, Cavenago Brianza, Bayer Diagnostici, Italy), glucose and serum creatinine by standard techniques, and blood pressure (BP) by mercury sphygmomanometer.

Simultaneous systolic BP determinations in the right forearm and both ankles were obtained through an 8 MHz linear Doppler probe (Stereodop, Promelec, Bourgeais, Guer, France). Ankle brachial index (ABI), an indicator of the haemodynamic severity of vascular disease [ 20], was expressed as the ratio of brachial to either tibial or pedal systolic pressure, whichever was higher [ 21]. ABI measurements were the values either at the leg with complaints or at the leg with the lowest value in the presence of bilateral disease.

Data analysis

Definition of categorical variables. Diabetes was any fasting venous blood glucose value above 140 mg dL−1, 2 h postload glucose > 200 mg dL−1 and/or current antidiabetic treatment. Hypertenson was defined as at least three systolic BP readings > 140 mmHg and/or diastolic BP > 90 mmHg, and/or current antihypertensive treatment. Total cholesterol and triglyceride status were dichotomized by taking values of 200 and 170 mg dL−1, respectively, as cut-offs. Smoking status was categorized as ‘current’, ‘former’ (having quit at least 1 year prior) and ‘never-smoker’.

Statistics

Descriptive statistics are expressed as mean ± standard deviation, or median (range) for skewed data. Statistical differences in the distribution of single characteristics were tested by univariate methods (unpaired two-tailed t-test and one-way analysis of variance for continuous variables, or χ2 statistics for discrete variables with Yates' correction). Pearson's correlation coefficients and standardized mortality rates per 100 patient-years were calculated by standard techniques.

To allow for differing follow-up intervals, univariate analysis of mortality was performed using the proportional hazard model to account for censored data and correlation with covariates. Independent predictors of death were identified by age-corrected, multivariate, stepwise Cox regression, introducing only those variables related univariately to outcome with a P-value below 0.05. Analysis did not include ABI values as, out of 108 patients, data were missing in 36, and not measurable in 17 due to arterial calcification. Relative risk (RR) (technically hazard ratios) and 95% confidence intervals (CI) were standardized by 1 SD for continuous values and to the bottom group for stratified parameters. The level of statistical significance was set at < 0.05.

Calculations were performed using Statgraphics Plus (Manugistic Inc., Release 1997, Rockville, MD, USA) and EGRET (SERC, © 1985–93, Seattle, WA, USA).

Results

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Demographic and clinical characteristics of the cohort are shown in Table 1.

Table 1.   Mean plasma lipoprotein concentration and mean processed plasma volume with LDL apheresis during concomitant simvastatin and atorvastatin therapy Table 1 Characteristics of the coeliac disease cases Table 1 Serum calcium levels in HIV patients according to CDC groups vs. controls. Calcium was adjusted for albumin, if it was below the lower reference of 2.20 mmol L−1 using Payne's equation (see }Methods'). Values are means SD. Table 1 Demographic and clinical variables (= 108). Data are median (range) or means ± SD Thumbnail image of

Females (± 30) were older (76 ± 7 vs. 70 ± 10 years, < 0.001) than males (= 78), with higher blood glucose levels [105 (65–338) vs. 93 (59–303) mg dL−1, < 0.01], greater prevalence of diabetes (56.7 vs. 30.7%, < 0.02), and a lower proportion of smokers (9 vs. 42%, < 0.001). Plasma fibrinogen levels did not differ significantly between the two sexes [females: 3.6 (2.2–8.8) vs. males: 4.1 (1.5–9) g L−1].

WCC (= 0.47, < 0.000 01) and serum creatinine (= 0.43, < 0.000 01) correlated positively, and haematocrit negatively (= −0.31, < 0.001), with fibrinogen. Fibrinogen was comparable between diabetics and non-diabetics [4 (1.5–9) vs. 3.8 (2.1–9) g L−1]. No variable showed a statistically significant relationship with age.

Median ABI was 0.3 (0–0.6, = 55) and was unrelated to plasma fibrinogen (= 0.02).

Mortality

Forty-five fatal events (36 cardiovascular, Table 2) were identified during a median follow-up of 1.6 years (8 days–5.5 years, for a total of 218 patient-years) ( Fig. 1a), with no differences in death rates between patients with rest pain and those with trophic lesions (9/27 vs. 36/81, respectively).

Table 2.   Details of events (= 45) Thumbnail image of
image

Figure 1.  Proportion of patients surviving at a given time during follow-up (Kaplan-Meier estimation method). The analysis shows time-to-death in the whole (n= 108, left panel), as well as in patients categorised by fibrinogen tertiles (cut-off points : 3.3, 4.5 g/L−1). See the main text for risk ratios. Legends report deaths in each category. The number of patients remaining alive and in-study at the beginning of each 12-month period is shown below the graphs.

Download figure to PowerPoint

The deceased tended to be older (75 ± 9 vs. 70 ± 10 years, < 0.006) with higher fibrinogen [4.4 (2.1–9.) vs. 3.3 (1.5–8.8) g L−1, < 0.001, and Fig. 1b for data categorized in tertiles], WCC [84 (25–178) vs. 76 (37–154) × 106 cells L−1, < 0.04] and serum creatinine [114.9 (53–389) vs. 97.2 (53–380) µmol L−1, < 0.04], and lower haematocrit (37.8 ± 4.4 vs. 40.2 ± 6%, < 0.01) at baseline. Other potential prognostic factors, such as diabetes and hypertension prevalence, smoking status, cholesterol, triglyceride and BP levels, were comparable. No significant difference was found between patients who died and those who survived with regard to the presence upon admission of coronary (40 vs. 24%) or cerebrovascular (17.7 vs. 14.3%) disease, or congestive heart failure (11.1 vs. 4.8%).

Twenty-three deaths (96% cardiovascular) occurred within 6 months, 11 (64% cardiovascular) within 2 years, and 11 (82% cardiovascular) after 2 years of follow-up ( Table 2); the corresponding standardized mortality rates were 340.7, 29.8 and 6.3 events (100 patient-years)−1. Plasma fibrinogen was the only baseline parameter (including age and others reported in Table 1) distinguishing (< 0.01) early (< 6 months) from later events ( Fig. 2).

image

Figure 2.  Plasma fibrinogen by time of death during follow-up. Box-and-whisker plot of baseline plasma fibrinogen distribution in patients dying within and after 6 months, as well as survivors at the end of the follow-up. Higher (< 0.01) plasma fibrinogen distinguishes early events from later ones and censored observations. The central box encloses the middle 50% of the data, and the median is plotted as a cross. Vertical lines (whiskers) extend from each end of the box and cover four interquartile ranges. Points identify outliers.

Download figure to PowerPoint

Fibrinogen as an independent predictor of death

Of the full set of baseline parameters ( Table 1), fibrinogen (< 0.001, standardized RR = 2.03, 95% CI = 1.75–2.34), serum creatinine (< 0.001, standardized RR = 1.15, 95% CI = 1.02–1.6), WCC (< 0.037, standardized RR = 3, 95% CI = 2.76–3.28) and haematocrit (< 0.04, standardized RR = 0.95, 95% CI = 0.89–0.99) were significantly related to survival in univariate age-corrected Cox regressions. Only fibrinogen was independently related to mortality (< 0.001, standardized RR = 2.03. 95% CI = 1.82–2.44) when the variables were subjected to multivariate age-corrected analysis according to the proportional hazard model. When fibrinogen distribution was divided in tertiles (cut-off points 3.3 and 4.5 g L−1), the risk of death approximately doubled with each tertile increase (2nd vs. 1st tertile: RR = 2.5, 95% CI = 1.1–5.6, < 0.03; 3rd vs. 1st tertile: RR = 4.1, 95% CI = 1.9–8.8, < 0.0001; Fig. 2, left panel).

Discussion

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The main and original finding of this longitudinal observation of elderly patients with chronic critical limb ischaemia is the association of elevated plasma fibrinogen with the relative risk of death. Mortality doubled for each standard deviation above the mean and increased in a graded manner with each tertile increase in fibrinogen, indicating that a specific threshold effect is lacking. The data strengthen fibrinogen's role as a predictor of mortality and are consistent with results obtained in stable claudicants [ 8, 10], geriatric patients [ 22], subjects with coronary artery disease [ 23] and even men [ 24, 25] and women [ 7] without obvious disease upon commencement of follow-up. Thus, our results confirm [ 16–18] the ominous implications for life expectancy of the appearance of resting pain and/or trophic lesions in previously uncomplicated peripheral arterial occlusive disease patients. However, protracted observation also showed lack of uniformity in the disease course of this group of critical limb ischaemia patients. In fact, an early period of exceedingly rapid accrual of mainly cardiovascular deaths was followed by a later phase in which demises occurred at a progressively slower rate and were less frequently due to cardiovascular causes. The reasons for such heterogeneous behaviour, unrecognized in previous follow-ups limited to 1 year or less [ 16–18], are unclear. In this regard, it is noteworthy that, of all the potential prognostic factors considered, elevated fibrinogen was the only parameter distinguishing the subset of patients with a more aggressive disease evolution. One might hypothesize that the existence of abnormal fibrinogen predated the development of critical limb ischaemia, perhaps as a consequence of a more advanced and progressive atherosclerotic process in these patients [ 6, 26]. However, this possibility is unlikely, firstly because fibrinogen levels were independent of ABI, a surrogate measure of systemic atherosclerosis [ 27] and haemodynamic severity of regional disease [ 21], and secondly, because the distribution of coronary, cerebrovascular and myocardial disease upon admission did not differ markedly between those who died during the following years and those who did not. Lack of differences linked to diabetes, age and sex is not surprising, as it concurs with findings in the general elderly population [ 28–30]. Interestingly enough, in the univariate analysis, fibrinogen showed a statistically significant positive association with serum creatinine, suggesting some link with worsening renal function. However, as no data are available for conditions less extreme than end-stage renal failure [ 31], we cannot speculate on this potentially interesting finding. Furthermore, other frequent correlates of fibrinogen, such as cholesterol, triglycerides, BP and smoking [ 28, 29], showed distributions independent of the outcome. Thus, the impact of those biological factors previously shown in both stable claudicants [ 4, 5] and healthy elderly populations [ 28–30] was probably overcome by stronger stimuli in our patients, particularly in the subgroup with the most progressive and unfavourable course of disease. Fibrinogen, an acute-phase protein [ 32], may respond to inflammatory stimuli triggered by the drastically reduced tissue oxygenation typical of chronic critical limb ischaemia [ 33], especially if complicated by sepsis and necrosis. Normalization of fibrinogen after successful vascular surgery of critically ischaemic limbs [ 34] implies that tissue ischaemia may indeed stimulate hepatic fibrinogen synthesis, possibly through interleukin-6 produced by activated monocytes [ 32]. The positive correlation with leucocyte levels is consistent with such a hypothesis, although the present evidence cannot be considered conclusive, as similar trends have been reported in elderly subjects unaffected by peripheral vascular disease [ 28, 29]. The influence of reactive components on fibrinogen is also consistent with the lower baseline haematocrit in those patients who died during the subsequent follow-up, as inflammatory processes depress haematopoiesis in a manner roughly proportional to the duration and severity of disease [ 35]. Similar inferences can be drawn from the tendency towards decreasing packed-cell volume with increasing fibrinogen values, an association opposite to that reported in uncomplicated subjects [ 36] and in contrast with conceptions of haematocrit as a prognostic predictor positively related to clinical events [ 36, 37]. However, even non-specific fibrinogen stimulation might still be a relevant chain in the link of events eventually leading to death, for example by compromising organ perfusion through increased blood viscosity [ 38] or by activating thrombosis [ 39]. On the other hand, the possibility that elevated fibrinogen levels may be a consequence, rather than a cause, of disease cannot be discounted. The nature of such an association will be clarified only by selective pharmacological lowering of fibrinogen.

In conclusion, plasma fibrinogen levels were found to be an independent predictor of early cardiovascular death in elderly peripheral arterial occlusive disease patients with critical limb ischaemia, in whom fibrinogen stimulation, perhaps related to inflammatory processes consequent to drastically impaired limb perfusion, may have contributed to the development of clinical events. However, further trials, possibly using specific fibrinogen inhibitors, are needed to confirm the precise mechanisms involved.

Acknowledgements

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The authors are grateful to Mr Lorenzo Evangelisti (II Clinica Medica, Università di Pisa) for conducting the fibrinogen assays. This paper was presented in preliminary form at the 6th Congress of the European Society of Atherosclerosis, 13–17 July 1996, Florence, Italy; the 46th Meeting of the American Society of Cardiology, 16–19 March 1997, Anaheim, CA, USA; the 19th Congress of the European Society of Cardiology, 24–28 August 1997, Stockolm, Sweden; and the 11th European Congress of Angiology, 23–26 October 1997, Rome, Italy.

References

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Dormandy JA, Hoare E, Colley J, Arrowsmith DE, Dormandy TL. Clinical, haemodynamic, rheological and biochemical findings in 126 patients with intermittent claudication. Br Med J 1973; 4: 576 81.
  • 2
    Hughson WG & Mann Garrod JI. Intermittent claudication: prevalence and risk factors. Br Med J 1978; i: 1379 81.
  • 3
    Stuart J, George AJ, Davies AJ, Aukland A, Hurlow RA. Haematological stress syndrome in atherosclerosis. J Clin Pathol 1981; 34: 464 7.
  • 4
    Lowe GDO, Fowkes FGR, Dawes J, Donnan PT, Lennie SE, Housley E. Blood viscosity, fibrinogen and activation of coagulation and leucocytes in peripheral arterial disease and the normal population in the Edinburgh Artery Study. Circulation 1993; 87: 1915 20.
  • 5
    Lee AJ, Lowe GDO, Woodward M, Tunstall-Pedoe H. Fibrinogen in relation to personal history of prevalent hypertension, diabetes, stroke, intermittent claudication, coronary heart disease, and family history: the Scottish Heart Health Study. Br Heart J 1993; 69: 338 42.
  • 6
    Lassila R, Peltonen S, Lepantalo M, Saarinen O, Kauhanen P, Manninen V. Severity of peripheral atherosclerosis is associated with fibrinogen and degradation of cross-linked fibrin. Arterioscler Thromb 1993; 13: 1738 42.
  • 7
    Kannel WB, D'Agostino RB, Belanger AJ. Update on fibrinogen as a cardiovascular risk factor. Ann Epidemiol 1992; 2: 457 66.
  • 8
    Fowkes FGR, Lowe GDO, Housley E, Rattray A, Rumley A, Elton RA, Mac Gregor IR, Dawes J. Cross-linked fibrin degration products, progression of peripheral arterial disease, and risk of coronary heart disease. Lancet 1993; 342: 84 6.
  • 9
    Wiseman S, Kenchington G, Dain R, Marshall CE, McCollum CN, Greenhalgh RM, Powell JT. Influence of smoking and plasma factors on patency of femoropoliteal vein grafts. Br Med J 1989; 299: 643 6.
  • 10
    Banerjee AK, Pearson J, Gilliland EL, Goss D, Lewis JD, Stirling Y, Meade TW. A six year prospective study of fibrinogen and other risk factors associated with mortality in stable claudicants. Thromb Hemostasis 1992; 68: 261 3.
  • 11
    Violi F, Criqui M, Longoni A, Castiglioni C. Relation between risk factors and cardiovascular complications in patients with peripheral vascular disease. Results from the A.D.E.P. study. Atherosclerosis 1996; 120: 25 35.
  • 12
    Ernst E & Resch KL. Fibrinogen as a cardiovascular risk factor: a meta-analysis and review of the literature. Ann Int Med 1993; 118: 956 63.
  • 13
    Singer A & Rob C. The fate of the claudicator. Br Med J 1960; ii: 633 6.
  • 14
    Dormandy JA & Murray GD. The fate of the claudicant – a prospective study of 1969 claudicants. Eur J Vasc Surg 1991; 5: 131 3.
  • 15
    European Working Group on Critical Leg Ischaemia Second European Consensus document on chronic critical leg ischaemia. Circulation 1991; 84 (supp IV): 1 26.
  • 16
    Wolfe JHN. Defining the outcome of critical ischaemia: a one year prospective study. Br J Surg 1986; 73: 321.
  • 17
    Lepantalo M & Matzke S. Outcome of unreconstructed chronic critical leg ischaemia. Eur J Vasc Endovasc Surg 1996; 11: 153 7.
  • 18
    The i.c.a.i. Group (gruppo di sudio delli'ischemia cronica critica degli arti inferiori). A prospective epidemiological survey of the natural history of chronic critical leg ischaemia. Eur J Vasc Endovasc Surg 1996; 11: 112 20.
  • 19
    Rossi E, Mondonico P, Lombardi A, Preda L. Method for the determination of functional (clottable) fibrinogen by the new family of ACL coagulometers. Thromb Res 1988; 52: 453 68.
  • 20
    Kiekara O, Riekkinen H, Soimakallio S, Lansimies E. Correlation of angiographically determined reduction of vascular lumen with lower-limb systolic pressures. Acta Chir Scand 1985; 151: 437 40.
  • 21
    Winsor T. Influence of arterial disease on the systolic blood gradients of the extremities. Am J Med Sci 1950; 220: 117 26.
  • 22
    Rumeau M & Pourrat JP. Prognostic value of fibrinogen level in geriatric patients. Lancet 1981; i: 951.
  • 23
    Benderly M, Graff E, Reicher-Reiss H, Behar S, Brunner D, Goldbourt U. Fibrinogen is a predictor of mortality in coronary heart disease patients. Arterioscler Thromb Vasc Biol 1996; 16: 351 6.
  • 24
    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; ii: 533 7.
  • 25
    Rosengren A & Wilhelmsem L. Fibrinogen, coronary heart disease and mortality from all causes in smokers and non smokers. The study of men born in 1933. J Int Med 1996; 239: 499 507.
  • 26
    ECAT Angina Pectoris Study GroupECAT angina pectoris study: baseline associations of hemostatic factors with extent of coronary atherosclerosis and other coronary risk factors in 3000 patients with angina pectoris undergoing coronary angiography. Eur Heart J 1993; 14: 8 17.
  • 27
    Mc Kenna M, Wolfson S, Kuller L. The ratio of ankle and arm arterial pressure as an independent predictor of mortality. Atherosclerosis 1991; 2–3: 119 28.
  • 28
    Cushman M, Yanez D, Psaty BM, Fried LP, Heiss G, Lee M, Polak JF, Savage PJ, Tracy RP. Association of fibrinogen and coagulation factors VII and VIII with cardiovascular risk factors in the elderly. The Cardiovascular Health Study. Am J Epidemiol 1996; 143: 665 76.
  • 29
    Sharp DS, Abbott RD, Burchfiel CM, Rodriguez BL, Tracy RP, Yano K, Curb DJ. Plasma fibrinogen and coronary heart disease in elderly Japanese-American men. Arterioscler Thromb Vasc Biol 1996; 16: 262 8.
  • 30
    Tracy RP, Bovill EG, Fried LP, Heiss G, Lee MH, Polak JF, Psaty BM, Savage PJ. The distribution of coagulation factors VII and VIII and fibrinogen in adults over 65 yrs. Results from the Cardiovascular Health Study. Ann Epidemiol 1992; 2: 509 19.
  • 31
    Bostom AG, Shemin D, Lapane KL, Sutherland P, Nadeau MR, Wilson PW, Yoburn D, Bausserman L, Tofler G, Jacques PF, Selhub J, Rosenberg IH. Hyperhomocysteinemia, hyperfibrinogenemia, and lipoprotein(a) excess in maintenance dialysis patients: a matched case control study. Atherosclerosis 1996; 125: 91 101.
  • 32
    Ernst E. Fibrinogen as a cardiovascular risk factor. Interrelationship with infections and inflammation. Eur Heart J 1993; 14 (Suppl. K): 82 7.
  • 33
    Lowe G. The biochemistry of critical limb ischaemia. Crit Ischaem 1991; 1: 11 23.
  • 34
    Woodburn KR, Rumley A, Lowe GDO, Pollock JG. Fibrinogen and markers of fibrinolysis and endothelial damage following resolution of critical limb ischaemia. Eur J Vasc Endovasc Surg 1995; 10: 272 8.
  • 35
    Erslev AJ. Anemia of chronic disease. In: Beutler E, Lichtman MA, Coller BS, Kipps TJ, eds. Williams Hematology, 5th edn. New York, NY. McGraw-Hill, 1995; 518 24.
  • 36
    Kannel WB, Wolf PA, Castelli WP, D'Agostino RB. Fibrinogen and risk of cardiovascular disease. The Framingham Study. JAMA 1987; 258: 1183 6.
  • 37
    Lowe GD, Lee AJ, Rumley A, Price JF, Fowkes FG. Blood viscosity and risk of cardiovascular events. The Edinburgh Artery study. Br J Haematol 1997; 96: 168 73.
  • 38
    Lowe GDO. Blood rheology in arterial disease. Clin Sci 1986; 71: 137 46.
  • 39
    Ceriello A, Pirisi M, Giacomello R, Stel G, Faletti E, Motz E, Lizzio S, Gonano F, Bartoli E. Fibrinogen plasma levels as a marker of thrombin activation: new insights on the role of fibrinogen as a cardiovascular risk factor. Thromb Haemost 1994; 71: 593 5.

Received 10 December 1997; accepted 15 May 1998.