Relationships between ovarian hormone concentrations and mental rotations performance in naturally-cycling women

Highlights

• Tested associations between ovarian hormones and mental rotations performance

• 528 naturally-cycling premenopausal women participated.

• Progesterone positively predicted mental rotations performance within subjects.

• No other relationships between estradiol or progesterone and MRT were found.

• Results should be interpreted cautiously due to risk of Type I error.

Abstract

Circulating gonadal hormones have been linked to variation in the structure and function of the adult human brain, raising the question of how cognition is affected by sex hormones in adulthood. The impacts of progestogens and estrogens are of special interest due to the widespread use of hormone supplementation. Multiple studies have analyzed relationships between ovarian hormones and mental rotation performance, one of the largest known cognitive sex differences; however, results are conflicting. These discrepancies are likely due in part to modest sample sizes and reliance on self-report measures to assess menstrual cycle phase. The present study aimed to clarify the impact of progestogens and estrogens on visuospatial cognition by relating mental rotation task performance to salivary hormone concentrations. Across two studies totaling 528 naturally-cycling premenopausal women, an internal meta-analysis suggested a small, positive effect of within-subjects changes in progesterone on MRT performance (estimate = 0.44, p = 0.014), though this result should be interpreted with caution given multiple statistical analyses. Between-subjects differences and within-subject changes in estradiol did not significantly predict MRT. These results shed light on the potential cognitive effects of endogenous and exogenous hormone action, and the proximate mechanisms modulating spatial cognition.

Introduction

Circulating gonadal hormones influence the structure and function of the adult primate brain in areas beyond those implicated directly in reproduction (Morrison, 2008). Mapping of estrogen and progesterone receptors in the brains of non-human primates and rodents has revealed receptor expression in regions such as the amygdala, hippocampus, cingulate cortex, midbrain raphe nuclei, central gray matter, and prefrontal cortex (McEwen, 2001). This distribution throughout the brain suggests that ovarian hormones may modulate cognitive sex differences, as well as changes in cognition across the menstrual cycle.

One of the largest known cognitive sex differences is the ability to visualize the rotation of an object around its axis (Miller and Halpern, 2014), with men outperforming women in mental rotation tests (MRT) by approximately 0.6 standard deviations (Jansen and Heil, 2010). A large body of research in humans and non-human animal models suggests that sex differences in spatial cognition result, at least in part, from organizing effects of androgen on the developing brain (Collaer and Hines, 2020; Puts et al., 2008; Shirazi et al., 2020). Although some studies have reported effects of testosterone administration on aspects of spatial cognition in small samples (Aleman et al., 2004; Pintzka et al., 2016), neither individual differences (Griksiene and Ruksenas, 2011; Puts et al., 2010; Shirazi et al., 2020; Wharton et al., 2008) nor intraindividual fluctuations (Griksiene et al., 2019; Puts et al., 2010; Silverman et al., 1999) in testosterone concentrations appear to predict mental rotations performance in healthy young adult women or men.

Some evidence suggests that MRT performance fluctuates across menstrual cycle phases (e.g., Hampson, 1990; but see e.g. Griksiene and Ruksenas, 2011), suggesting modulation by progesterone and/or estradiol (reviewed in Poromaa and Gingnell, 2014). However, studies exploring this possibility present conflicting results (Griksiene and Ruksenas, 2011; Hampson et al., 2014; Noreika et al., 2014; Zhu et al., 2015), and differ in whether the design is between-subjects (e.g., Griksiene et al., 2019), within-subjects (e.g. Zhu et al., 2015), or both (e.g., Courvoisier et al., 2013). The few studies that have simultaneously evaluated between- and within-subjects effects have involved small samples (e.g., Courvoisier et al., 2013, n women = 7), which cannot provide reliable estimates of population effects (Gangestad et al., 2016).

Due to methodological limitations in previous investigations and the considerable variability in past results, we utilized data from two previously completed studies to investigate both interindividual and intraindividual effects of progesterone and estradiol on MRT. Specifically, in our combined sample of 528 premenopausal, naturally-cycling women, we tested whether average ovarian hormone levels predicted differences in MRT between subjects, and whether intraindividual fluctuations in these hormones predicted changes in MRT across two study sessions.