Freedivers' brain networks reorganize over seven months of breath-hold training

Exercise on its own is consistently associated with reorganization of large-scale brain networks ^[9], and seven months of freediving training is also seven months of exercise. Compared against sedentary controls, any change in the divers could be put down to fitness. Recruiting controls whose aerobic load matched the divers' ^[5] means a difference between the groups points at the breath-holding instead. The first thing to check in the preprint is whether those controls were also scanned before and after. ScienceAlert's description of the two scans mentions only the freedivers ^[6].

The hippocampus is where trouble would be expected. It helps form and retain memories, and its high energy demands make it particularly sensitive to hypoxia ^[10]. Oxygen deprivation has been found to interfere with neurogenesis and synaptic plasticity, two processes that learning and memory depend on ^[11]. Yet earlier studies of freedivers found none of the cognitive impairment associated with clinical hypoxia ^[12].

The scans covered breath-holds as well as normal breathing. The apnea rounds add up to at most eight minutes of breath-holding per session ^[1], inside a block of up to 14 minutes once recovery breathing is counted ^[2]. That let the team compare connectivity during and after each breath-hold, then compare both across the seven months ^[7]. The memory test let them ask whether those connectivity changes went along with how well the divers' memories were working ^[13].

Micaux and her colleagues wrote that the findings "suggest that repeated voluntary hypoxia during freediving training is associated with selective functional reorganization of hippocampal and large-scale brain networks" ^[14]. "This pattern may reflect adaptive neuroplasticity linked to preserved episodic memory under intermittent hypoxic exposure," they wrote ^[15]. The work is a bioRxiv preprint and has not yet been peer reviewed ^[3].

I think the authors chose verbs that fit the design. One group of volunteer divers, scanned twice, can show that network connectivity changed while memory held up. It cannot show that the change is what kept memory intact. Even a reorganization that tracked memory scores diver by diver would still be a correlation. The adaptation idea goes back to the team's 2025 hippocampus result, which suggested to the researchers that something adaptive may be taking place ^[16]. This study fits that idea. Testing whether the network changes cause the preserved memory would take a different design.