2 - (SYM 6) Unique Associations of Suicide Attempt History (vs. Ideation Only) with Premenstrual Changes in Emotional Symptoms, Suicidal Ideation, and Suicide Planning

Recent studies reveal high lifetime rates of suicide attempt among patients with premenstrual dysphoric disorder, and many cross-sectional studies demonstrate hospitalizations for suicide attempt occur most commonly around menses. However, it remains unclear whether there is a unique association between suicide attempts and premenstrual mood changes, over and above associations with depression and suicidal ideation. In a preregistered analysis of 175 naturally-cycling patients (18-45 yrs) recruited for past-month suicidal ideation, we hypothesized that a history of suicide attempt (vs. ideation only) would be associated with greater premenstrual suicide and emotional symptom changes. Suicide history was measured with the C-SSRS or SITBI-R. P</span>articipants provided daily ratings of psychosocial suicide-related symptoms (i.e., depression, mood lability, anxiety, anger, perceived burdensomeness, belonging, passive ideation, active ideation, and planning thoughts) and menstrual cycle status for at least one month. We used multilevel growth spline models (time coded as –13 to +12 around menses, with a knot at menses onset) with days (level 1) nested within participants (level 2). We examined the interaction between suicide attempt history and time (premenstrual and postmenstrual splines) predicting daily symptoms, controlling for age, SSRI use, and their interaction with time variables. Results indicated individuals with an attempt history (vs. Ideation only) reported overall greater mean levels of mood liability, passive ideation, active ideation, perceived burdensome, and planning thoughts (estimates 0.10-0.36, p’s < 0.05). Although history of attempt (vs. ideation only) was not associated with degree of premenstrual change in emotional symptoms or passive ideation (estimates –20-0.26, p’s > 0.5), it did predict a more rapid increase in active ideation premenstrually (estimate 0.06, p = 0.001) and a slower recovery post-menstrually (estimate -0.03, p = 0.047). In addition, attempt history (vs. Ideation only) predicted a more rapid increase in the likelihood of suicidal planning premenstrually (OR = 1.56, p = 0.095) and belonging (estimate –0.18, p = 0.09). Together, these results reinforce a unique association between suicide attempt (vs. ideation only) and premenstrual changes in active suicidal ideation, planning, and belonging. This analysis supports further investigation into potential causal effects of the menstrual cycle not only on suicidal ideation severity, but also on acute risk for suicide attempt and death.