Changes in Visual Long-Term Potentiation Show Preserved Cyclicity in Human Females Taking Combined Oral Contraceptives

Abstract

Introduction: The combined oral contraceptive (COC) pill is often employed to address physical and neurological symptoms in menstrual cycle-related disorders by suppressing shifts in endogenous gonadal hormone fluctuations. Symptom persistence, especially in the lead up to the hormone-free interval (HFI), suggests an underlying neurobiological mechanism of preserved cycling. Our study utilised a non-invasive method of visually inducing long-term potentiation (LTP) to index changes in neural plasticity in the absence of hormonal fluctuations.

Methods: Visually induced LTP was recorded using electroencephalography in 24 healthy female COC users across three sessions: days 3 and 21 during active hormone pills, and day 24 during the HFI. The Daily Record of the Severity of Problems (DRSP) questionnaire tracked premenstrual symptoms. Dynamic causal modelling (DCM) was used to elucidate the neural connectivity and receptor activity changes associated with LTP across different days of COC.

Results: Visually induced LTP was greater on day 21 than day 3 (p = 0.011) and was localised to the P2 visually evoked potential. There was no effect of the HFI (day 24) on LTP. DCM of differences between days 3 and 21 showed changes to inhibitory interneuronal gating of LTP in cortical layer VI. The DRSP only showed a significant increase in symptoms in the HFI, meaning the LTP result appeared more sensitive to cyclicity.

Conclusions: This study provides objective evidence of preserved cyclicity in COC users through enhanced LTP on day 21 compared to day 3 of a 28-day COC regimen, indicating that relatively higher excitation in the brain despite peripheral gonadal suppression may underlie and exacerbate menstrual cycle-related disorders.

Keywords: Combined oral contraceptive pill; Dynamic causal modelling; Electroencephalography; Gonadal hormones; Long-term potentiation.

Fig. 1.

Evoked potential analysis pipeline. The analysis pipeline from the initial ANOVA that determines the follow-on analyses (one-way ANOVAs) that test each hypothesis. a An effect of the combined oral contraceptive (COC) on long-term potentiation (LTP) compared to the HFI. b An effect of cyclicity while on COC.

Fig. 2.

Diagram illustrating the timing and structure of conditions for the visual LTP task. Adapted from Sumner, McMillan .

Fig. 3.

a, b Six-cell thalamocortical model architecture and connectivity parameters. Cortical populations include layer II/II with superficial pyramidal (sp), and superficial inhibitory interneuron (si) input. Layer IV has ss input and is also the source of thalamic rl input as well as sp input. Layer V is populated by deep pyramidal (dp) and deep inhibitory interneuron (di) input. Layer VI is populated with the thalamic pyramidal (TP) cells, the source of rl input also. The model also parameterises the decay constant of AMPA, NMDA, GABA-A, GABA-B, and M and H channels. The checkered/dashed parameters were fixed, and the solid parameters were allowed to vary with task effects.

Fig. 4.

DRSP global scores. Mean of the total (global) score of the DRSP on each day. Error bars represent standard error of the mean.

Fig. 5.

DRSP physical symptoms. Percent incidence of a DRSP symptom (score >1) for physical symptoms over each day.

Fig. 6.

a Oestradiol and progesterone levels measured at each session. Bar graphs show the mean concentrations, participant 5’s day 3 progesterone was removed from this graph as it was an outlier that substantially affected the interpretation of the group (33.4-fold greater than the next closest value). Error bars represent standard error of the mean. b The mean study concentrations of progesterone and oestradiol plotted alongside data from the literature on the unmedicated menstrual cycle. Progesterone and oestradiol concentration from the study (days 3, 21, and 24). Reference ranges are taken from Verdonk, Vesper for oestradiol, and the comparably timed mean concentrations of progesterone from (tabled data presented in [53]). Ranges are taken from the late follicular phase (approximately days 9–13, comparable to day 3 of COC), the late-luteal phase (approximately days 26–28, comparable to day 21), and the early follicular phase (approximately days 1–8, comparable to day 24).

Fig. 7.

Raw VEPs on day 24, used to test for the effect of LTP. a N1 shown at representative electrode (P7). b P2 is shown at representative electrode (POz). The time window selected for subsequent analyses is shaded in orange. c Topography of early post-tetanus minus pre-tetanus localising the effect of tetanising speed on visual LTP to the bilateral P1.

Fig. 8.

a Topographies at 174 ms. Topography of the P2 at 174 ms (peak of the effect of cyclicity on LTP). b Mean extracted field peak at 174 ms. c Individual peaks extracted at 174 ms from the field. d Late post-tetanus minus pre-tetanus difference waves at representative electrode for the P2 Oz and electrode proximal to the field peak PO9. Note: Though only days 3 and 21 were compared statistically, day 24 is included in panels a, b, and d for illustrative purposes. Online supplementary Figure S1 shows Fig. 8c grouped by hormonal composition of COC taken by participants.

Fig. 9.

Modulation of parameters for visual LTP, day 3 versus day 21. Parametric empirical Bayes (PEB) results demonstrating reduced changes to ss > si, si > ss, and tp > rl connectivity in the linear model of visual LTP on day 21 compared to day 3.