Facial Emotion Recognition and Emotional Memory From the Ovarian-Hormone Perspective: A Systematic Review

We review original papers on ovarian-hormone status in two areas of emotional processing: facial emotion recognition and emotional memory. Ovarian-hormone status is operationalized by the levels of the steroid sex hormones 17β-estradiol (E2) and progesterone (P4), fluctuating over the natural menstrual cycle and suppressed under oral contraceptive (OCs) use. We extend previous reviews addressing single areas of emotional processing. Moreover, we systematically examine the role of stimulus features such as emotion type or stimulus valence and aim at elucidating factors that reconcile the inconsistent results.

We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and included papers published until September 2020 indexed in PubMed and Web of Science databases. Search terms were MeSH terms (emotional OR emotion) AND (X) AND (estrogen OR progesterone OR menstrual cycle OR oral contraceptives) with (X) representing our separately searched areas, resulting in (processing OR recognition OR empathy), and (memory OR recall). To be included, articles had to (1) be written and published in English, (2) examine healthy, non-pregnant adult women in their reproductive age, and (3) measure or at least estimate levels of E2 and P4. In PubMed, the search was (4) limited to humans and (5) to the search term present in the title or abstract.

Features of the provided stimulus material (emotion type and/or valence) constitute a relevant influence that interacts with E2- and P4-related ovarian-hormone status. For instance, recognition of basic emotions appears to be more related to P4- than E2-levels. Quite consistent, OC intake (vs. natural menstrual cycling) was accompanied by impaired recognition accuracy of basic and also complex emotions, although not in a recent large-sample study assessing complex emotions. Memory recall of negative content was mainly enhanced by P4, especially after having been stressed.

We document the methodological diversity in the field, presumably contributing to the heterogeneity of results. More studies explicitly contrasting the early follicular phase, mid-cycle phase, mid-luteal, and OC intake while standardizing tasks are needed. Research would take advantage of using within-subject designs and accounting for the recognition of complex emotions.