Preliminary results: NREM sleep alterations, hormonal shifts and overnight memory consolidation during early pregnancy

Introduction:
Many pregnant women experience sleep disruptions and memory problems during pregnancy; the latter is also known as „Baby Brain“. Nevertheless, the existence of objectively measurable memory loss is still debated, and studies examining this phenomenon are rare. Our ongoing project investigates how memory-critical electrophysiological features of NREM sleep, particularly fast and slow sleep spindles, as well as slow oscillations, are altered and influenced by sex hormones (estradiol and progesterone) during early pregnancy.
Method:
Seven pregnant women (Mage=29.29±1.89) in their first trimester and 12 non-pregnant women (Mage=25.62±2.90) in their luteal phase participated in a seven-day experiment. Subjects had to fill out several questionnaires (e.g., Pittsburgh Sleep Quality Index) and polysomnography was recorded during an adaptation and a learning-night. Progesterone and estradiol were assessed via saliva samples. A declarative word-pair task was performed before and after the learning night. Sleep architecture, slow (11-13 Hz) and fast (13-15 Hz) sleep spindles, slow oscillations, and their coupling were analyzed.
Results:
Pregnant women showed comparable overnight memory consolidation to non-pregnant women (U=31.00, p=.373, r=0.26). Subjective sleep quality (U=5.00, p=.008, r=0.77) and slow spindle density (U=11.00, p=.048, r=0.63) were significantly reduced, while wake-time-after-sleep-onset (U=1.00, p=.001, r=0.96) was significantly increased in pregnant compared to non-pregnant women. The percentage of N3 sleep, fast spindle density and the positive peak amplitudes of slow oscillations were not altered during pregnancy. Progesterone was negatively correlated with fast spindle density in non-pregnant women (rs=-0.63, p=.042). Furthermore, higher progesterone and estradiol were related to lower peak amplitudes of slow oscillations in non-pregnant women, whereas in pregnant women, the opposite was found: higher progesterone and estradiol were associated with higher peak amplitudes.
Conclusion:
Our preliminary results indicate that pregnant women exhibit alterations in memory-related NREM sleep parameters as early as the first trimester. Some of these parameters were related to sex hormones, but hormonal alterations did not show memory-enhancing nor impairing effects in this sample. However, these results need further confirmation by examining the final sample (N=20 per group). Furthermore, our currently ongoing investigation of women across pregnancy will allow elegant within-subject analyses and thereby interesting insights into the development of these changes over pregnancy.