Table 2.
Age-adjusted, sex-adjusted and race-adjusted associations of coagulation factor IX and coagulation factor XI antigen levels with cardiovascular disease risk factors in REGARDS at baseline
| Characteristic | FIX (%) |
FXI (%) |
||
|---|---|---|---|---|
| Difference in level (95% CI) | P-value | Difference in level (95% CI) | P-value | |
|
| ||||
| Age (per 9.3 years) | 1.3 (0.33–2.3) | 0.009 | − 0.4 (− 1.6 to 0.75) | 0.49 |
| Female sex | 8.2 (5.6–10.7) | < 0.0001 | 13.0 (10.1–15.8) | < 0.0001 |
| Black race | 2.0 (− 0.5 to 4.5) | 0.11 | − 1.8 (− 4.6 to 1.1) | 0.22 |
| Stroke belt region | 2.2 (−0.4, 4.7) | 0.10 | −3.1 (−6.0, −0.2) | 0.04 |
| Systolic BP (per 16 mmHg) | 3.5 (2.3–4.8) | < 0.0001 | 0.04 (− 1.3 to 1.4) | 0.95 |
| Antihypertensive medication use | 9.1 (6.4–11.8) | < 0.0001 | 3.0 (− 0.04 to 6.1) | 0.05 |
| Total cholesterol (per 39 mg dL−1) | 3.6 (2.1–5.0) | < 0.0001 | 3.6 (1.9–5.2) | < 0.0001 |
| HDL cholesterol (per 17 mg dL−1) | − 2.3 (− 4.0 to − 0.60) | 0.008 | − 0.62 (− 2.4 to 1.1) | 0.48 |
| Lipid-lowering medication use | 7.1 (4.4–9.8) | < 0.0001 | 9.0 (5.9–12.0) | < 0.0001 |
| Never smoked | 0.0 (Reference) | – | 0.0 (Reference) | – |
| Former smoker | 2.9 (0.11–5.6) | 0.04 | 2.5 (− 0.49 to 5.5) | 0.10 |
| Current smoker | 6.4 (2.5–10.3) | 0.001 | 3.5 (− 1.4 to 8.4) | 0.16 |
| BMI (per 5.9 kg m−2) | 5.2 (4.0–6.4) | < 0.0001 | 3.1 (1.7–4.5) | < 0.0001 |
| Diabetes | 7.2 (4.1–10.3) | < 0.0001 | 8.7 (4.6–12.7) | < 0.0001 |
| Atrial fibrillation | − 5.0 (− 10.6 to 0.6) | 0.08 | − 0.02 (− 5.1 to 5.1) | 0.99 |
| LVH | 6.7 (2.7–10.7) | 0.001 | 0.08 (− 4.7 to 4.9) | 0.97 |
| CRP (per 1.17) | 8.5 (7.3–9.7) | < 0.0001 | 5.1 (3.7–6.6) | < 0.0001 |
| D-dimer (per 0.87) | 5.3 (3.8–6.9) | < 0.0001 | 4.6 (2.9–6.2) | < 0.0001 |
| FIX antigen (per 22%) | – | – | 9.1 (7.5–10.8) | < 0.0001 |
| FXI antigen (per 25%) | 8.1 (6.7–9.4) | < 0.0001 | – | – |
BMI, body mass index; BP, blood pressure; CI, confidence interval; CRP, C-reactive protein; LVH, left ventricular hypertrophy. Differences in coagulation factor levels were estimated by the use of β-coefficients evaluated with linear regression models per standard deviation higher value (shown in parentheses) adjusted for age, sex, and race. In models of age, sex, and race, the estimates were adjusted only for the two remaining variables. CRP and D-dimer were natural log-transformed.