174 Chapter 8 Table 2. Linear regression model examining the association of migraine diagnosis, BMI, and age with 1) peak estradiol plasma levels (pmol/L) during the luteal phase (T8) and 2) the luteal estradiol decline (pmol/L) between day 8 and 14. Results presented as log2-transformed estimates, standard error (SE), and corresponding back-transformed estimates (2Estimate) with 95% confidence intervals (CI). Estimate SE 2Estimate 95% CI p-value 1) Peak estradiol (pmol/L) luteal phase (day 8) Migraine diagnosis 0·49 0·21 1·49 1·03 – 1·87 0·04 BMI 0·04 0·03 1·03 0·03 – 0·04 0·21 Age -0·01 0·02 0·34 0·01 – 0·02 0·74 2) Estradiol decline (pmol/L) luteal phase (day 8-14) Migraine diagnosis 0·93 0·36 1·90 1·16 – 3·12 0·03 BMI 0·17 0·06 1·12 1·03 – 1·22 0·01 Age -0·03 0·03 0·98 0·95 – 1·02 0·29 Secondary outcomes Pathway analysis showed differences in peak estradiol levels (p = 0·04) and estradiol decline (p = 0·02) during the luteal phase (days 8–14) between women with migraine and healthy controls (Table 2, Figure 2). Figure 3 visualizes variables included in the pathway analysis, showing their correlation with either migraine participants or controls and corresponding p-values. Visual inspection suggested a higher testosterone-to-estradiol (T/E2) ratio on day 8 (mid-luteal phase) in individuals with migraine. However, this difference was no longer present after adjusting for BMI and age in the pathway analysis (p = 0·24). Similarly, no differences were found in other hormonal subsets, including the group of estradiol, progesterone, LH, and FSH (p = 0·70), SHBG (p = 0·81) or androgens (p = 0·72). Mean plasma levels with 95% confidence intervals for progesterone, testosterone, FSH, and LH are shown in Figure 4. Sensitivity analyses Excluding women aged ≥45 resulted in a mean age of 31 ± 5·8 years in the control group (n = 18) and 35 ± 6·2 years in the migraine group (n = 24; p = 0·01). Peak plasma estradiol levels were 38% higher in the migraine group (2Estimate = 1·38, 95% CI: 1·05–1·87, p = 0·03), adjusted for BMI and age. A second sensitivity analysis, excluding women aged ≥40, showed no significant age difference (migraine group: 33·7 ± 5·5 years, n = 20; control group: 30·5 ± 5·8 years, n = 18; p = 0·09). Peak plasma estradiol remained 38% higher in the migraine group (2Estimate = 1·38, 95% CI: 1·03–1·95, p = 0·04), adjusted for BMI and age.
RkJQdWJsaXNoZXIy MjY0ODMw