Smoking, other risk factors and fibrinogen levels: Evidence of effect modification

PURPOSE: Recent investigations have indicated that smoking may act as an important modifier of the risk associated with dyslipidemia. We hypothesized as a potential mechanism that smoking modifies the association between traditional risk factors of early atherosclerosis, such as dyslipidemia, hypertension, or diabetes, with fibrinogen, a risk factor more closely related to plaque progression and thrombosis. METHODS: The sample for this cross-sectional study, was collected from the MONICA Augsburg population survey elf 1989/90 and included 1840 men and 1784 women, aged 25 to 64 years. Traditional risk factors hypertension, diabetes mellitus, and dyslipidemia were assessed by personal interviews, and medical examinations. Plasma fibrinogen concentration was determined by nephelometry. Effect modification was assessed by stratified analyses and tests for statistical interaction. RESULTS: Fibrinogen levels were higher in women than in men and, after adjustment for potential confounders, higher in smokers than in nonsmokers (each p < 0.001). The effect of smoking was greater in men (p < 0.001). The elevation of mean adjusted fibrinogen levels associated with hypertension or with diabetes in men was significantly higher in smokers as compared to nonsmokers (test for interaction, p < 0.001). By contrast, smoking in women showed significantly stronger impacts only with regard to the association of dyslipidemia and fibrinogen (p < 0.001). Comparing groups of subjects with increasing numbers of concomitant risk factors, stratified according to their smoking status, effect modification by smoking was particularly evident in male participants with two or more risk factors. Numbers of women with more than one risk factor and smoking were too low for analysis. CONCLUSIONS: We demonstrate that smoking contributes more than additively to the strong influences of single and combined traditional risk factors on fibrinogen levels. These data confirm that smoking is a dominant determinant of fibrinogen levels in the general population. Ann Epidemiol 2001;11:232-238. (C) 2001 Elsevier Science Inc. All rights reserved.