Table 3.

Four-way decomposition of the effect of smoking on stroke by venules

	Current smokers compared to non-smokers
	Estimate (95% CI)
Total effect	1.41 (1.10; 1.67)
Excess relative risk due to each component
Controlled direct effect	0.32 (0.03; 0.58)
Reference interaction	0.02 (− 0.03; 0.03)
Mediated interaction	0.06 (− 0.00; 0.11)
Pure indirect effect	0.02 (− 0.01; 0.04)
Proportion of effect due to each component
Controlled direct effect	77% (60%; 120%)
Reference interaction	4% (− 10%; 8%)
Mediated interaction	15% (− 10%; 28%)
Pure indirect effect	4% (− 8%; 11%)
Overall proportion
Mediated	19% (− 12%; 31%)
Attributable to interaction	19% (− 18%; 35%)
Eliminated	22% (− 20%; 39%)

Values are adjusted for age, sex, subcohort, education, systolic blood pressure, diastolic blood pressure, blood pressure lowering medication use, body mass index, total cholesterol, high-density lipoprotein cholesterol, white blood cell counts, diabetes mellitus, alcohol intake, carotid plaques, history of cardiovascular disease, and retinal arteriolar caliber