Cortisol dynamics and sleep quality: the role of sex and oral contraceptive use 123 could indicate that the effects of OC on sleep disruptions could be dependent on CAR, but this hypothesis needs further study in prospective setups. In men, diurnal levels of testosterone and cortisol in men are strongly correlated (Harden et al., 2016), meaning testosterone could affect cortisol dynamics. Vice versa, sleep deprivation can affect testosterone production, resulting in lower testosterone production the next morning (Su et al., 2021). This may implicate that poorer sleep could possibly reduce testosterone levels, which could be accompanied by a reduction in cortisol dynamics. Methodological considerations The main strengths of the current study lie in the methodological rigor. We preregistered the data selection and analysis, and in a part of our aims, we aimed to replicate previous findings. Participants were screened for previous or current psychiatric disorders, meaning that our results are less likely to be confounded by possible differences in psychiatric history between groups. We aimed to replicate study findings by Morssinkhof et al. (2021), and we used a similar item to the original study and state-of-the-art analyses to study these ordinal scores. Furthermore, the CAR was sampled using five measurements in the first 60 minutes after awakening, which increases the validity of the CAR sampling (Nasser et al., 2023). However, the following limitations should be taken into account when interpreting our study. Firstly, OC users in this study were using different formulations and generations of OCs, as shown in Supplementary Figure S4.1 in the Supplementary materials. Previous studies found that effects of OCs on cortisol dynamics and sleep could be dependent on OC formulation (Herrera et al., 2019; Partonen et al., 2023), and future studies should also account for OC formulations. Secondly, as shown in Supplementary Table S4.1 in the supplementary materials, most of the NC women in our sample were in the follicular phase at the time of the measurements, and participants in the luteal phase were underrepresented. Both sleep quality and cortisol dynamics could differ in different menstrual phases (Baker & Driver, 2007; Wolfram et al., 2011), so future studies should determine whether menstrual phase could affect the association between sleep and CAR. Thirdly, this study only used ordinal scores from a single item on subjective sleep disruptions in the PSQI, which is a less valid method than
RkJQdWJsaXNoZXIy MTk4NDMw