Twelve healthy older adults climbed into 40°C water for one hour at a time, repeated the routine for six weeks and then came back with a mixed result. Some tests of working memory and logical reasoning improved, but the study also reported several important non-changes: sleep stayed the same, blood markers tied to Alzheimer disease stayed the same and the main planned memory outcome did not improve.
The Experimental Physiology study came from researchers led by the University of Portsmouth. It was a small, uncontrolled pre-post intervention, which means the same participants were measured before and after the six-week program without a separate comparison group. That design can show whether signals are worth pursuing, but it cannot prove that hot water immersion alone caused every change.
How the six-week hot water plan worked
Participants averaged about 69 years old and the group included four men and eight women. Each person completed two to three weekly sessions of hot water immersion, with a minimum of 12 and a maximum of 18 baths across six weeks. During each session, rectal temperature was held between 38.5 and 39.0°C, so the intervention delivered a real heat load rather than a short warm soak.
Researchers tested thinking during the first and final immersion sessions at three points: before entering the water, immediately after leaving it and three hours later. The cognitive battery covered 1-back and 2-back tasks for working memory, a 2-choice reaction-time task for inhibition and speed and a logical relations task for reasoning. Those tasks were run through the ANAM platform, a neuropsychological test system whose repeat-testing characteristics have been studied before, which matters because repeated exposure can itself influence scores.
Outside the task sessions, the team added broader measures to see whether a cognitive shift matched a physical one. They measured cerebral oxygenation during testing with near-infrared spectroscopy, checked resting common carotid artery blood flow before and after the intervention, tracked one week of sleep with a wrist accelerometer before the first baths and during the final week and measured plasma amyloid-β42 plus phosphorylated tau before and after the six weeks.
Which thinking tests actually improved
The best results came from the easier working-memory task and the reasoning task, not from every measure in the battery. Repeated immersion improved 1-back performance and logical relations performance across the intervention. When the authors compared grand means, logical relations rose from 22.48 to 25.31 responses per minute and 1-back rose from 67.92 to 73.96 responses per minute in the participants with usable data for those analyses.
The paper treated 2-back throughput immediately after immersion as the primary outcome and that result did not improve. Scores were 53 responses per minute during the first hot-water session and 55 during the final session, a difference that was not statistically significant. The 2-choice reaction-time measure also showed no main effect, so the study did not deliver a broad cognitive lift across memory, speed, inhibition and reasoning all at once.
That pattern is the main reason to keep cognition at the center of the story while staying cautious. The study suggests that repeated heat exposure may help some mental operations in older adults, especially lighter working memory demands and logical reasoning. At the same time, the unchanged primary outcome means the most important planned memory test did not move, so the article cannot honestly frame the intervention as a blanket memory boost.
Researchers also tried to reduce one obvious alternative explanation by familiarizing participants with the tasks before the intervention began. The paper reports no differences between the penultimate and final familiarization sessions for any of the four tests. That helps, but it does not erase the possibility of lingering practice effects, day-to-day variation, motivation differences, or simple random movement in a group this small.
Why the physiology results complicate the mechanism
A popular idea in this field is that passive heat might sharpen cognition by improving blood delivery or oxygen handling in the brain. A 2020 review on heat therapy and the aging brain outlined several candidate pathways, including vascular changes, heat-shock responses and downstream effects on brain health. The new study tested part of that story directly instead of leaving it as theory.
The immediate physiology results went in an awkward direction for a simple boost narrative. A single hot-water session significantly reduced all four measured cerebral oxygenation variables right after immersion. Three hours later those values had returned to baseline, which shows an acute disruption followed by recovery rather than a straightforward rise that tracked with better same-day thinking.
The six-week comparison was even more revealing. Repeated immersion produced no chronic adaptation in those oxygenation measures and the study found no significant correlations between cerebral oxygenation and any of the four cognitive tasks immediately after immersion in either the first or final hot-water session. Resting carotid blood flow also did not change significantly from before the intervention to after it.
Those null physiology findings do not cancel the cognitive improvements that did appear, but they do narrow what can be claimed. If hot water immersion helped some parts of cognition here, the mechanism was not captured by chronic changes in the measured oxygenation variables or resting common carotid artery flow. The paper itself ends by calling for more work on mechanism, which is a more defensible conclusion than assuming the vascular hypothesis was confirmed.
What stayed unchanged after six weeks
Sleep was one of the clearest non-results. Using a wrist-worn method based on earlier accelerometer sleep work described in PLoS ONE, the researchers found no meaningful change in sleep efficiency, sleep duration, or time in bed between the baseline week and the final intervention week. That null result also sits awkwardly beside an older systematic review that suggested passive body heating before sleep may help sleep quality in older adults, which makes the unchanged sleep data here more notable rather than less.
The biomarker results were also flat. Plasma amyloid-β42 and phosphorylated tau, two blood measures linked to Alzheimer pathology research, did not change after six weeks of immersion. Two participants had unreadable amyloid-β42 values, so that analysis used 10 people, while the phosphorylated tau analysis used all 12.
Those unchanged outcomes deserve as much attention as the improved reasoning and 1-back scores because they define the boundary of the evidence. The study did not show that passive heating altered sleep quality, altered resting carotid flow, altered chronic cerebral oxygenation during testing, or shifted these blood biomarkers of neurodegeneration over six weeks. Readers should not come away thinking that a hot-water routine has already been shown to modify Alzheimer related biology in healthy older adults.
The same caution applies to the phrase neurodegenerative biomarkers in the paper title. In this experiment, the researchers measured circulating markers associated with disease research, not clinical dementia outcomes, brain scans, or long-term diagnosis. A six-week intervention with 12 participants is far too small and short to settle whether passive heat can influence the course of neurodegenerative disease.
Why the design needs a careful reading
The most important limit is the one built into the protocol: this was a small, uncontrolled pre-post study. Everyone received the hot-water intervention, no one served as a no-treatment or sham comparison group and some analyses used data from 8, 9, or 10 people rather than all 12 because of missing or unusable measurements. That structure makes the work useful for hypothesis generation, but it leaves plenty of room for other explanations.
Sample size matters even more when the result pattern is selective. Logical reasoning improved, 1-back improved, 2-back did not improve, 2-choice reaction time did not improve, sleep did not improve and the biomarker measures did not improve. A larger randomized trial would be needed to show whether the gains in some tasks were reproducible, whether they survive a true control comparison and whether the unchanged outcomes stay unchanged in a broader population.
The study still contributes something valuable. It shows that repeated passive heat in older adults is feasible under controlled conditions and that cognition can move in a domain-specific way rather than a simple all-or-nothing direction. That is a more interesting result than a marketing-style hot bath claim, because it suggests future trials should ask which tasks are most sensitive, how long any benefit lasts and whether a different schedule or temperature might change the parts of the picture that remained fixed here.
For now, the strongest reading is narrow and evidence-based. Repeated hot water immersion was followed by better performance on one reasoning task and one lighter working-memory task in a very small group of older adults, while the primary memory outcome, reaction-time measure, sleep metrics, carotid flow, cerebral oxygenation adaptation, amyloid-β42 and phosphorylated tau all stayed unchanged. That combination makes the study worth following, but it does not justify claims that hot baths broadly improve cognition or alter Alzheimer related biology.
