Embodied memory for emotional faces: Effects of cardiac cycle phases and post-learning stress on identity and expression memory.

Human learning and memory are considered 'embodied', but it is unclear which body signals affect which memory facet. We investigated whether ascending cardiac signals affect memory encoding and consolidation. Due to its relevance for social behavior, this study investigated identity and expression memory for emotional faces. To implement different conditions of ascending cardiac signals for encoding, participants viewed happy and angry faces for 150 ms time-locked either to the early (R-wave +230 ms) or late cardiac cycle phase (R-wave +530 ms), which differ in their feedback from arterial baroreceptors. Participants then underwent either a stress (n = 30; Socially Evaluated Cold Pressor Test/SECPT) or a no-stress procedure (n = 29) to apply conditions of high vs. low cardiac signals in the consolidation phase. Recognition of face identity and emotional expression was tested 30 min and 24 h after acquisition. Memory for both identity and expression was significantly better when encoded in the late cardiac cycle phase, both for happy and angry faces. Memory performance declined over time. Although the SECPT increased heart rate, systolic (SBP) and diastolic blood pressure, as well as salivary cortisol, it did not affect memory. Nevertheless, stress responsiveness of cortisol was a positive predictor and SBP responsiveness was a negative predictor of expression memory. Our findings suggest that increased baroreceptor feedback globally attenuates declarative memory encoding for emotional faces (i.e. identity/expression memory), independent of valence. In contrast, the effect of perturbations in ascending cardiac signals by acute stress on consolidation of identity or expression memory depends on stress axis responsiveness.