Scans across three memory stages caught older adults merging two events into one

The useful part of this design is the 20 minutes in which participants did nothing. Brain activity kept being recorded through that gap, between the moment each face-image pair was studied and the test that followed ^[5]. Recording it lets a team ask which phase an error enters, instead of scoring only the final answer. The hippocampus is active at all three stages, and the firing patterns from the initial experience often recur while a memory is being stored and again when it is retrieved ^[12].

Older adults were less likely to repeat those patterns across the three stages ^[8]. Then the second finding, which is the awkward one: when older adults did repeat the original patterns, they were even more likely to confuse their memories, a result the authors call surprising ^[9]. In young adults, reinstatement of the original pattern is treated as the signature of a memory being replayed as recorded ^[12]. In the older group, the same signature did not tell you whether the answer was right. The authors' interpretation is that the hippocampus is inappropriately changing memories between memory phases as a person gets older ^[10].

One of the study's authors, writing in The Conversation, wrote that the work "shows how the brain blends related experiences as you age rather than replaying the exact one you need" ^[2].

The test itself records more than a score. Each trial put a previously seen face above four images, and participants had to say which one had been paired with it ^[6]. With four options, chance is 25 percent ^[18]. So an error arrives with information attached, namely which wrong image was chosen. Older adults' mistakes ran toward related events, falsely remembering a face first paired with an apple as having been paired with a forest ^[7]. Scientists already knew that confusing memories happens more often than forgetting them altogether ^[11]. A test scored only right or wrong cannot separate an event that was never stored from two events that merged.

The Conversation article does not report how many people were scanned or how large the age differences were ^[19]. The comparison is across young, middle-aged and older groups measured at one time, so it shows differences between age groups. It does not show memory changing inside a person over years ^[20]. And the material being remembered is a face paired with a forest or an apple, imagined for a few seconds in a scanner ^[4]. The conversation-with-a-friend example is the article's illustration, not the stimulus.

On what the brain does when an older adult gets a memory right, the authors are direct. "So what is the brain signature when older adults get a memory right? Are other brain regions outside the hippocampus helping?" the author wrote in The Conversation, adding, "We did not get a good answer for these questions from our study" ^[13][14]. The article names electrical stimulation of the hippocampus as one of the next frontiers in neuroscience, in the sense of understanding how it prevents memory loss or restores lost memories ^[15]. Future interventions might try to strengthen the hippocampus to stop it altering a memory after storage ^[16].