Approximately 6% of ovulating individuals experience the debilitating affective symptoms of Premenstrual Dysphoric Disorder (PMDD), which arise premenstrually and remit post-menses [1]. Meanwhile, nearly 60% of female patients with depression experience cyclical worsening similar to PMDD but without full post-menstrual remission, termed Perimenstrual Exacerbation (PME) [2]. These are both clinically-significant manifestations of Menstrual Cycle Affective Change (MCAC), a term encompassing the full range of variation in nature, severity, and timing of affective responses to the menstrual cycle. Given its recurring biological trigger, MCAC offers a unique opportunity to develop mechanistic biomarkers; below, we argue that this work can be strengthened through greater attention to temporal, mechanistic, and phenotypic specificity.

Second, more precise coding and modeling of dynamic biomarker variables are needed for testing specific hormonal mechanisms (i.e., mechanistic specificity). Given the dynamic nature of MCAC, biomarker development requires that raw measurements be transformed to correspond to one’s specific hormonal hypothesis. For example, the hypothesis that higher-than-usual levels of progesterone provoke symptoms requires person-centering (i.e., subtracting the person mean), whereas hypotheses about absolute hormone levels or changes require biomarkers coded as raw levels or change/slope scores, respectively [5]. Greater consideration of temporal lags is also needed.