Does recruitment into ripple activity depend on connection strength formed during wakefulness, explaining why only some experiences are replayed?

Ripples are the brain's replay timing signal

A ripple is a very brief burst of fast, synchronized electrical activity. In rodents it runs at roughly 150–250 cycles per second and lasts a fraction of a second [1]; in human intracranial recordings the comparable events are about 70 milliseconds long at around 90 cycles per second [3]. These bursts are most common in the hippocampus during sleep, and they coordinate activity across many brain regions at once [2].

During a ripple, the hippocampus replays a sequence of locations or experiences. In rodents, place cells — neurons that fire when an animal is at a particular spot — fire during replay at progressively earlier phases of each ripple cycle, sweeping forward along the remembered path [1]. That is a timing code: which neuron fires, and exactly when within the ripple, carries the trajectory. So ripples are not just noise; they are a structured format for re-running an experience.

Which neurons get pulled into a ripple?

The clearest human evidence comes from direct recordings of 1,466 neurons in the medial temporal lobe during learning, wakefulness, and sleep. Ripples reliably drove neurons to fire, and sleep ripples produced stronger activation than wake ripples. Crucially, neurons tuned to items that were later remembered fired more strongly during ripples than neurons tuned to items that were forgotten — and this memory-linked reactivation appeared only during sleep [4]. That is a selective recruitment pattern: not every experience gets equal replay, and the ones that stick are the ones whose neurons join in more strongly.

Ripples also seem to recruit partners across the brain rather than acting locally. In humans, cortical ripples co-occur and phase-lock with each other across widely separated sites, even between hemispheres, and this synchrony strengthens as more cortical sites join in [3]. Cortical ripples co-occur with hippocampal ripples but do not phase-lock with them, which the authors take to suggest that cortico-cortical synchrony is carried by direct cortico-cortical connections rather than by the hippocampus conducting the orchestra [3]. During waking, these hippocampal-cortical and cortical-cortical co-ripples increase just before a successful delayed memory recall, when a cue has to be bound to its response [3]. In a working-memory task, ripple activity varied systematically in a way that could predict cortical connection strength and distance [5].

Does wake-time connection strength decide who joins?

This is where the evidence stops short of your hypothesis. The studies describe what ripples do and which neurons participate, but none of them measures how strongly connections were formed during wakefulness and then tests whether that strength predicts recruitment into replay. The rodent phase-precession work is about the timing code of replay, not about which synapses were strengthened beforehand [1]. The review of sleep ripples ties them to memory consolidation and synaptic plasticity in general terms [2]. The human cortical work points to cortico-cortical connections as the route for synchrony, but it does not measure their wake-time strength as a predictor of replay [3].

The closest thing to a link is indirect. The finding that remembered items' neurons fire more during sleep ripples [4] is consistent with the idea that something about those experiences made their neurons more likely to be recruited — but the study does not identify that something as connection strength formed during wakefulness. Likewise, the observation that ripple activity can predict cortical connection strength and distance [5] is about ripples relating to connection properties, not about wake-time learning setting those properties and thereby selecting replay. So the honest answer is: the selective recruitment is real, but the specific causal chain you propose has not been demonstrated in these studies.

Why sleep replay may matter more than waking replay

One consistent theme is that sleep is a privileged time for this process. Sleep ripples drove stronger neuronal activation than wake ripples, and the memory-linked reactivation — remembered items firing more than forgotten ones — was seen only during sleep [4]. Ripples are also most frequently observed in the hippocampus during sleep, where they coordinate hippocampus, neocortex, and subcortical areas [2].

That does not mean waking ripples are irrelevant. In humans, cortical ripples occur during both sleep and waking, and waking co-ripples rise before successful recall [3]. The difference may be one of degree and context: waking ripples appear to support binding and retrieval in the moment, while sleep ripples appear to drive the stronger, selective reactivation associated with consolidation [4]. What remains open is whether the selectivity seen in sleep is set up by connection strength laid down during the preceding wake period — a plausible idea, but one these studies do not directly test.

Sources used in this answer

Related research questions