Table 2

 Risk factors associated with ischaemic stroke: results of multivariable logistic regression

Compared with community controlsCompared with hospital controls
Odds ratio (95% CI)p ValueOdds ratio (95% CI)p Value
HDL, high density lipoprotein.
All models adjusted for age and sex. Model A also adjusts for antihypertensive treatment, lipid or blood sugar lowering treatment.
OR here should be per 1 SD increment in fasting blood sugar, systolic blood pressure, triglycerides, total and HDL cholesterol (1 SD for cases; see table 1).
All ischaemic strokes
    (A) Models with individual risk factors
        Fasting blood sugar4.55 (1.63, 12.67)0.0041.17 (0.68, 2.02)0.57
        Total cholesterol0.86 (0.44, 1.68)0.661.92 (0.97, 3.82)0.06
        HDL cholesterol0.17 (0.09, 0.30)<0.0010.27 (0.17, 0.44)<0.001
        Triglycerides0.70 (0.37, 1.32)0.270.67 (0.40, 1.11)0.12
        Smoking7.77 (1.93, 31.27)0.0043.95 (1.61, 9.71)0.003
        Systolic blood pressure1.88 (1.01, 3.49)0.0451.13 (0.74, 1.73)0.57
    (B) Models with clustering of metabolic syndrome components
        ⩾3 components4.76 (1.93, 11.76)<0.0012.09 (1.06, 4.13)0.03
Strokes excluding cardioembolic and strokes due to other aetiologies
    (A) Models with individual risk factors
        Fasting blood sugar4.33 (1.45, 12.96)0.0091.12 (0.58, 2.19)0.73
        Total cholesterol1.10 (0.51, 2.39)0.802.45 (1.09, 5.51)0.03
        HDL cholesterol0.15 (0.07, 0.30)<0.0010.26 (0.15, 0.46)<0.001
        Triglycerides0.77 (0.39, 1.50)0.440.77 (0.42, 1.42)0.41
        Smoking8.73 (2.14, 35.68)0.0035.38 (2.01, 14.39)0.001
        Systolic blood pressure2.35 (1.11, 4.96)0.031.29 (0.79, 2.12)0.31
    (B) Models with clustering of metabolic syndrome components
        ⩾3 components5.75 (2.24, 14.76)<0.0012.38 (1.16, 4.87)0.02