Abstract
Despite decades of research, there is no scientific consensus method for representing the menstrual cycle as a continuous timeline. Common phase- and count-based methods oversimplify hormonal dynamics and overlook individual variability in ovulation timing, reducing statistical power and misaligning trajectories. To address this, we introduce Phase-Aligned Cycle Time Scaling (PACTS) and its companion R package, `menstrualcycleR`, which generates continuous time variables anchored to both menses and ovulation, improving alignment of hormonal dynamics across individuals and cycles in an accessible, reproducible way. This approach accommodates variable cycle lengths and supports various ovulation detection methods or norm-based ovulation estimation when biomarkers are unavailable. Using daily urinary hormone data from 44 cycles, we show that PACTS improves alignment and estimation of estradiol (E2) and progesterone (P4) trajectories compared to traditional methods. This has implications for improved cycle effect estimation both within samples and across studies. The variables produced by PACTS also support hierarchical nonlinear models, such as generalized additive mixed models, for high-resolution analysis of cyclical outcomes and quantification of related effect sizes. Since a variety of clinical disorders are currently diagnosed using cycle count methods, future directions should consider how this approach may improve diagnostic accuracy (e.g., in premenstrual dysphoric disorder (PMDD), catamenial epilepsy, or menstrual migraine). Together, PACTS and `menstrualcycleR` offer a reproducible framework that improves precision and interpretability in menstrual cycle research.