Abstract
BACKGROUND: Some women who use cyclic hormonal contraception (CHC) suffer from premenstrual symptoms; whether their symptoms differ from women who do not use CHC is not clear. OBJECTIVE: To compare women who use or do not use CHC on perimenstrual symptom timing and change severity. STUDY DESIGN: We analyzed daily symptom ratings from women who requested participation in (Screened Cohort: 103 used CHC and 387 did not) or were randomized in (Randomized Cohort: 41 used CHC and 211 did not) a clinical trial for premenstrual syndrome. We used effect sizes to compute and compare change scores between cycle phases in four partially overlapping perimenstrual windows defined relative to day 1 of menses [(-6, -1), (-5, 1), (-4, 2), (-3, 3)]. Differences in magnitude of change and timing were estimated using linear mixed-effects models. RESULTS: Both cohorts showed a significant two-way interaction between CHC use and symptom change scores (p < 0.01) and a significant main effect of perimenstrual window (p < 0.0001). Overall menstrual cycle symptom change was greater for the nonhormonal contraception versus hormonal contraception group. In the Screened Cohort, change scores were greater in the nonhormonal group specifically for depression (p = 0.04); anger or irritability (p < 0.01); and physical symptoms (p < 0.01). Mean change scores increased as the window shifted forward toward menses for both cohorts with the largest effect size and greatest group difference for (-4, 2) interval. CONCLUSIONS: CHC slightly attenuates menstrual cycle symptom change. The (-4, 2) perimenstrual interval shows the largest change compared with postmenses.