Hemostatic factors and risk of coronary heart disease in general populations: new prospective study and updated meta-analyses

Abstract

Background: Activation of blood coagulation and fibrinolysis may be associated with increased risk of coronary heart disease. We aimed to assess associations of circulating tissue plasminogen activator (t-PA) antigen, D-dimer and von Willebrand factor (VWF) with coronary heart disease risk.

Design: Prospective case-control study, systematic review and meta-analyses.

Methods: Measurements were made in 1925 people who had a first-ever nonfatal myocardial infarction or died of coronary heart disease during follow-up (median 19.4 years) and in 3616 controls nested within the prospective population-based Reykjavik Study.

Results: Age and sex-adjusted odds ratios for coronary heart disease per 1 standard deviation higher baseline level were 1.25 (1.18, 1.33) for t-PA antigen, 1.01 (0.95, 1.07) for D-dimer and 1.11 (1.05, 1.18) for VWF. After additional adjustment for conventional cardiovascular risk factors, corresponding odds ratios were 1.07 (0.99, 1.14) for t-PA antigen, 1.06 (1.00, 1.13) for D-dimer and 1.08 (1.02, 1.15) for VWF. When combined with the results from previous prospective studies in a random-effects meta-analysis, overall adjusted odds ratios were 1.13 (1.06, 1.21) for t-PA antigen (13 studies, 5494 cases), 1.23 (1.16, 1.32) with D-dimer (18 studies, 6799 cases) and 1.16 (1.10, 1.22) with VWF (15 studies, 6556 cases).

Conclusions: Concentrations of t-PA antigen, D-dimer and VWF may be more modestly associated with first-ever CHD events than previously reported. More detailed analysis is required to clarify whether these markers are causal risk factors or simply correlates of coronary heart disease.

Figure 1. Study flow diagram of the updated meta-analyses.

The figure was designed based on the 2009 PRISMA flow diagram template (available from http://www.prisma-statement.org/statement.htm).

Figure 2. Associations of baseline t-PA antigen, D-dimer and VWF with coronary heart disease risk (Reykjavik Study).

Odds ratios (95% CI) for coronary heart disease are shown by fifths of baseline t-PA antigen, D-dimer and VWF, plotted against the geometric mean level in each category on a log-doubling scale. *Adjusted for age, sex, period of recruitment, smoking status, body mass index, systolic blood pressure, history of diabetes, total cholesterol and log_e triglycerides.

Figure 3. Head-to-head comparison of associations of various baseline variables with coronary heart disease risk (Reykjavik Study).

*Values were log_e transformed for analysis. †Adjusted for age, sex, period of recruitment, smoking status, body mass index, systolic blood pressure, history of diabetes, total cholesterol and log_e triglycerides.

Figure 4. Meta-analyses of reported associations of t-PA antigen, D-dimer and VWF with coronary heart disease risk in prospective population-based studies.

Study acronyms are explained in the legend of Table 3 . Summary estimates were calculated using random effects models. *Degree of adjustment:+minimally adjusted (typically adjusted for age and sex only);++plus adjustment for at least one non-lipid marker;+++plus adjustment for at least one lipid marker;++++plus adjustment for at least one inflammatory marker. Where studies reported relative risks with more than one level of statistical adjustment, the most adjusted estimate was used (least adjusted estimates are reported in Figure S8).