Effects of menstrual cycle phase on associations between the error-related negativity and checking symptoms in women

Abstract

The menstrual cycle is known to impact mood and cognitive function and has been shown to lead to variability in symptoms of obsessive-compulsive disorders and anxiety. Using a within-subject design, the present study examined ovarian hormones, the error-related negativity (ERN), and self-reported checking symptoms in both the mid-follicular and mid-luteal phases of the menstrual cycle. ERN amplitude and checking symptom severity did not vary between the follicular and luteal phases. However, a more negative ERN was associated with greater checking symptoms in the luteal phase of the menstrual cycle, even when controlling for ERN amplitude in the follicular phase. Moreover, changes in checking symptoms between phases were associated with phase-related changes in the ERN. Finally, a significant mediation model was found such that the ERN measured in the luteal phase mediated the association between progesterone in the luteal phase and checking symptoms in the luteal phase. Collectively, the present findings suggest that levels of progesterone in the luteal phase could impact checking symptoms by modulating response monitoring and sensitivity to errors, and that fluctuation in the ERN between menstrual cycle phases may play an important role in the expression of anxious and obsessive-compulsive symptoms.

Keywords: Anxiety; Checking; EEG/ERP; Error-related negativity; Menstrual cycle; OCD; Progesterone.

Figure 1.

Schematic depicting the natural variability in the menstrual hormones estradiol and progesterone across the menstrual cycle. Reprinted with permission from Mulligan et al., 2018, Psychophysiology, e13268.

Figure 2.

Response-locked ERPs (left) for error and correct trials measured during the mid-luteal phase, and corresponding topographic maps for the correct-error difference (middle) in individuals low (top) and high (bottom) in checking symptoms. In the mid-luteal phase, individuals with high checking symptoms showed a more negative ΔERN (i.e., the difference in amplitude between correct and error conditions) as compared to individuals with low checking symptoms. The scatter plot (right) depicts the association between ΔERN and checking symptoms in the mid-luteal phase and the line represents the line of best fit.

Figure 3.

Scatter plot depicting the association between change in ΔERN from the mid-follicular to the mid-luteal phase, and change in checking symptoms from the mid-follicular to the mid-luteal phase. The line represents the line of best fit.

Figure 4.

Schematic depicting the a significant mediation model in which the ΔERN measured in the mid-luteal phase mediates an indirect association between progesterone in the mid-luteal phase and checking symptoms in the mid-luteal phase.