110 Chapter 5 Table 3. Results of mixed logistic regression model with migraine attack without aura as dependent variable. The nonsignificant interaction term was dropped in the second model, providing interpretable estimates for the perimenstrual window and migraine subtype. Estimate SE Test-statistic p-value OR (95% CI) Estimate SE Test-statistic p-value OR (95% CI) Intercept −2.52 0.03 −74.96 <0.001 0.08 (0.08–0.09) −2.51 0.03 −75.74 <0.001 0.08 (0.08–0.09) Perimenstrual window (ref = outside window) 0.45 0.04 11.46 <0.001 1.57 (1.45–1.70) 0.42 0.03 13.37 <0.001 1.53 (1.44–1.62) Migraine diagnosis (ref = MO) −0.36 0.05 −6.54 <0.001 0.70 (0.63–0.78) −0.38 0.05 −7.12 <0.001 0.69 (0.62–0.76) Perimenstrual window * migraine diagnosis −0.08 0.07 −1.21 0.224 0.92 (0.81–1.05) MO, migraine without aura; SE, standard error; OR, odds ratio; CI, confidence interval. Table 4. Results of mixed logistic regression model with migraine attack with aura as dependent variable. Only women with MA were included (n = 210). Estimate SE Test-statistic p-value OR (95% CI) Intercept −4.19 0.14 −30.16 <0.001 0.02 (0.01–0.02) Perimenstrual window (ref = outside window) 0.08 0.08 1.01 0.314 1.08 (0.93–1.26) MA, migraine with aura; SE, standard error; OR, odds ratio; CI, confidence interval.
RkJQdWJsaXNoZXIy MjY0ODMw