Why do some memories fade while others endure?

Ah, the age-old question: why do we remember what we remember? One possible mechanism is selective “memory replay” during sleep, in which the brain reactivates specific patterns of neuronal firing as seen during learning. In other words, memories that are rehearsed during sleep will most likely be retained (“consolidated” in neurojargon) in the long run. Studies in rats imply that not all memories are equal: periods involving reward information or crucial choice-points – “important” memories – are preferentially replayed and consolidated. The same seems to happen in humans – you probably remember your “emotionally salient” first kiss much better than what you had for breakfast that morning. Is replay responsible for this selectivity?  Does replay happen when you’re asleep or awake? Can we manipulate replay to remember what would’ve been forgotten?

Delphine Oudietete et al. (2013) The Role of Memory Reactivation during Wakefulness and Sleep in Determining Which Memories Endure. J. Neurosci. 33(15): 6672-6678.

To answer these questions, researchers worked with 60 young volunteers, and asked them to remember the location of objects on a screen (see figure below). Each object had a value displayed with it to inform volunteers how much money they could make if they recalled the location correctly. A distinctive sound was played for each object (“meow” for cat and “woof” for dog) – as you will see, these sounds are the puppet strings used later to manipulate memory replay.

After learning, participants were separated into four groups. Groups 1& 2 went for a 90min power nap. For group 2, researchers played sounds associated with half of the low-value objects during a phase of sleep called slow-wave sleep (e.g. researchers played only "meow" even though both cat (associated) and dog (not associated) are both low-value objects). The idea is those sounds will act as cues to trigger replay of associated low-value memories. Group 1 was left alone to snooze in peace. Groups 3 & 4 stayed awake. Group 3 watched a relaxing nature documentary while group 4 worked through a bunch of difficult memory tasks. During these tasks, researchers played sounds characteristic of half of the low-value objects in the background, similar to group 2. Because the task was so taxing, researchers assumed group 4 would not be able to pay attention to the sounds. In a sense, this design mimics the unconscious perception of sounds during sleep.

Testing the power-nappers for recall showed that while memory performance decreased in general, high-value object locations were remembered better than low-valued ones after sleep. However, playing sounds associated with low-value objects during sleep “rescued” the memory, to the extent that memory retention was similar for high- and low-valued objects. This seemed to work best when the sounds were played during slow-wave sleep: sound cues presented in another stage of sleep weren’t nearly as helpful. Surprisingly, sound cues helped the participants remember the whole category of low-value object locations better, not just the half directly associated with the sounds. Using our cat/dog example as above, playing "meow" during sleep enhanced memory retainment for both cat AND dog.

A similar trend showed up in the waking groups, with participants showing less error in remembering high-value object locations upon recall. While sound cues also enhanced retention of low-value object memories, further analysis showed that this was specific to the ones directly associated with the sounds, not the entire category.

So what makes us remember what we remember? Well, memories are selectively retained during both wakefulness and sleep, depending on their value. This selectivity seems to be driven by memory reactivation. Memories with a high (monetary) value are (probably) preferentially replayed and consolidated, as seen by the fact that high-valued objects show less memory decline. It’s more than likely that the same advantage holds for other types of salient (striking) memories. In fact, some researchers propose that disrupting replay and consolidation during sleep can weaken traumatic memories and reduce the chance of developing post-traumatic stress disorders.

Conversely, low-value memories can be saved from forgetting by artificially driving their reactivation through an associated cue. Presumably, this is because cues can cause reactivation of associated memories, increasing their replay and subsequent consolidation. Unfortunately, without sticking electrodes into the brain to record neural activity, we can’t be sure the sound cues in this study are triggering memory replay.

Which makes me wonder: What if you give cues corresponding to high-value objects during sleep and wakefulness? Will this strengthen the memory even more? Or will it eventually start interfering with the memory? Does the characteristic of the cue matter? Here it seems that the cues inherently reflect the nature of the object. What if a "baaahhh" is paired with a dog? Would this still strengthen the dog-related low-value memory? Why do sound cues enhance memory retention of the entire category of low-value objects when given during sleep, but increases only cue-associated memories during wakefulness? Can we specifically enhance individual memories during sleep? What happens if the cue is given throughout the entire sleep cycle, instead of just slow-wave sleep?

Finally, this study suggests we may be able to "sleep hack" our way into remembering several different types of memories better.

As shown above, memories can be explicit (facts, visual memory, abstract knowledge) or implicit (certain motor skills you can do without thinking). A recent study in Nature Neuroscience showed that replaying a newly learned piano melody during sleep helps people perform the melody better upon awakening, demonstrating that sound cue-induced reactivation can enhance motor (implicit) memories.This study suggests that explicit memories may also be open to sleep hacking. While I don't want to jump to conclusions, maybe next time I'll try falling asleep to my study playlist and see what happens.

Oudiette D, Antony JW, Creery JD, & Paller KA (2013). The Role of Memory Reactivation during Wakefulness and Sleep in Determining Which Memories Endure. The Journal of neuroscience : the official journal of the Society for Neuroscience, 33 (15), 6672-8 PMID: 23575863