# Eighteen expert saturation divers descended to a simulated 440 meters and became slower plus less accurate when conflicting numbers tested their attention at 45 atmospheres

> Expert divers working under pressure equal to 440 meters of seawater took longer to judge conflicting numbers and made more mistakes, according to a chamber experiment in Japan. The results suggest that extreme pressure can weaken parts of mental performance even when...

Canonical URL: https://www.argo.net/eighteen-expert-saturation-divers-descended-to-a-simulated-440-meters-and-became-slower-plus-less-accurate-when-conflicting-numbers-tested-their-attention-at-45-atmospheres/
Byline: ARGO.net Editorial Team
Published: 2026-08-18T05:40:02+00:00
Categories: Explainer, Oceans

![Concept of human intelligence with human brain on black background](https://www.argo.net/wp-content/uploads/2026/08/brain_medical_illustration-2.jpg)

Expert divers working under pressure equal to 440 meters of seawater took longer to judge conflicting numbers and made more mistakes, according to a chamber experiment in Japan. The results suggest that extreme pressure can weaken parts of mental performance even when divers are highly trained and breathe a helium-based gas designed for deep operations.

The [peer-reviewed study](https://doi.org/10.1186/s40101-024-00366-3), published in the **Journal of Physiological Anthropology**, examined [18 male saturation divers](https://pubmed.ncbi.nlm.nih.gov/39375772/) from the Japan Maritime Self-Defense Force. Researchers compared their performance at normal surface pressure with results recorded at 45 atmospheres absolute, or 45 ATA, inside a [deep-diving simulator](https://pmc.ncbi.nlm.nih.gov/articles/PMC11459827/).

The divers remained healthy and completed their training without reported neurological problems. Yet a carefully designed number test detected slower responses and reduced accuracy when the numbers sent conflicting signals. Such small changes could become important during deep missions, where a diver may need to follow instructions while handling equipment under strict time limits.

## A dive chamber recreated 440 meters of seawater

**Saturation diving** allows people to live under raised pressure for extended periods. After enough time, their body tissues hold as much dissolved breathing gas as the surrounding pressure allows. Remaining at depth then adds little extra gas, so divers can complete long work periods and undergo one slow decompression at the end.

The experiment took place at the **JMSDF Undersea Medical Center** in Yokosuka, Japan. Its deep-diving simulator includes a living chamber, a connecting lock and a wet chamber for underwater training. The cognitive tests were conducted in the dry living chamber, where researchers could control pressure and gas conditions closely.

At the surface, pressure is about 1 ATA. Each additional 10 meters of seawater adds roughly one atmosphere, bringing the simulated 440-meter depth to 45 ATA. The divers were gradually compressed to the target pressure and later passed through a slow decompression schedule that extended across the three-week diving operation.

The chamber contained a **heliox breathing mixture** made from helium and oxygen. Its temperature stayed between 30 and 32 degrees Celsius, or 86 to almost 90 degrees Fahrenheit, because dense helium carries heat away from the body efficiently. Humidity remained near 50 to 60 percent, while carbon dioxide and carbon monoxide were kept at low levels.

## Number conflicts tested mental control

Researchers measured cognition with **numerical Stroop tasks**. A Stroop test creates a conflict between two features of the same object, forcing the brain to focus on the feature that answers the question while holding back an automatic response.

For the numerical task, divers saw two digits and selected the one with the greater value. The physical size of each printed digit could support the correct choice or point toward the wrong one. A small-looking 7 beside a large-looking 2, for example, required the participant to ignore appearance and choose 7.

The physical task reversed the instruction. Divers selected the digit printed in the larger type while ignoring its numerical value. A large 2 beside a small 7 created a conflict because the brain quickly recognizes 7 as the greater number even though its printed size is smaller.

Each task block contained 288 trials, including experimental trials and filler trials that prevented the pattern from becoming too predictable. Researchers measured both **reaction time** and the percentage of correct answers. Half of the divers completed their surface-pressure comparison before the saturation dive, while the other half completed it afterward, which helped limit learning effects.

## Reaction times slowed at 45 ATA

At 45 ATA, divers needed more time to answer the conflicting trials in the numerical task. Their response times for matching and neutral number pairs remained statistically similar to their surface results, which suggests that the strongest slowdown appeared when physical size interfered with numerical meaning.

The physical comparison task showed a broader effect. Responses at 45 ATA were slower for matching pairs, neutral pairs and conflicting pairs. The delays were usually measured in hundredths of a second, yet the pattern appeared across the different conditions and reached statistical significance.

Researchers interpret Stroop interference as a test of **executive control**, the mental process that guides attention and suppresses distracting information. The results indicate that extreme hyperbaric exposure placed an added load on this control system, especially when the correct response had to overcome a competing signal.

A small delay during a computer test does not predict a specific accident. The study did not examine tool handling, emergency decisions or performance in open water. Its findings identify a measurable change in laboratory-style cognition that could help researchers design more direct tests of work performance at great depth.

## Conflicting choices produced more errors

Accuracy also fell under extreme pressure. In both number tasks, the **correct rate** decreased at 45 ATA when the printed size and numerical value pointed toward different answers. Accuracy in matching and neutral trials remained broadly stable.

The difficult close-number trials offer a concrete example. In one numerical condition, average accuracy fell from about 87 percent at surface pressure to about 77 percent at 45 ATA. Greater numerical separation made the answer easier, since the difference between 2 and 9 is more obvious than the difference between 1 and 2.

The combination of slower responses and lower accuracy gives researchers more evidence than either measurement alone. A person can sometimes preserve accuracy by working more slowly. During the conflicting trials, the divers gained less protection from extra response time, since their error rate also increased.

The paper's abstract states, "Our findings suggest that divers' cognition is impaired during 45 ATA deep SD." The authors applied that conclusion to performance on the Stroop tests. All 18 divers still completed the broader training operation without reported health problems or disruptions to the diving drill.

## Pressure effects remained with heliox

Deep divers often breathe helium mixed with oxygen because helium has far less narcotic action than nitrogen. At moderate depths, nitrogen narcosis can cloud thought, slow reactions and alter judgment. Helium allows divers to reach much greater depths while reducing that particular hazard.

Heliox still places the body in an unusual physical environment. Gas becomes much denser as pressure rises, breathing resistance increases and heat moves away from the body more quickly. At very high pressures, divers may also face **high-pressure nervous syndrome**, which can involve tremors and changes in brain activity.

The JMSDF compression plan was arranged to reduce the risk of high-pressure nervous syndrome. Testing at 45 ATA occurred after the divers had time to adjust and none showed obvious symptoms during the experiment. Earlier research has found that brain measurements can detect changes associated with [nitrogen narcosis](https://doi.org/10.1038/s41598-022-08869-8), while the present experiment examined a much deeper heliox environment.

Helium's low narcotic effect leaves several possible explanations for the slower thinking. Pressure itself may influence nerve activity. The dense gas may increase physical effort and living in a confined chamber may add mental strain. The experiment measured their combined effect during the saturation operation, so it cannot assign the result to one cause.

## The causes remain difficult to separate

The chamber protected the divers from many hazards of a real dive, including cold seawater and demanding underwater movement. The dry environment therefore allowed a cleaner look at deep pressure, while still including the confinement and long schedule of saturation diving.

Several limits narrow the conclusions. Every participant was an experienced male military diver, with an average age of about 37. Their training and physical condition may have reduced the effects. Results could differ among less experienced workers or a more varied group of divers.

Researchers compared only 1 ATA and 45 ATA. The safety schedule left too little time for full testing at intermediate pressures such as 31 or 35 ATA. As a result, the study cannot show where the decline began or whether performance changed steadily as pressure increased.

The team also lacked direct recordings of brain activity. According to authors **Nozomu Kageyama and Takehito Sawamura**, future work could combine cognitive testing with EEG or event-related potential measurements. A broad [diving cognition review](https://doi.org/10.3390/biology12020229) has likewise described how results can vary with depth, breathing gas, task design and the conditions surrounding each dive.

Professional divers may also underreport discomfort because they are motivated to complete an operation. Daily health checks found no major problems, although subtle physical strain could have influenced performance. Isolation and difficult communication may have contributed as the divers remained inside the enclosed system.

## Training could reduce deep-diving risks

The researchers reported that the findings influenced JMSDF saturation-diving education. Divers are taught that attention and mental control may change at extreme depth, even when they feel healthy and remain able to carry out routine duties.

Awareness can support safer work practices. Teams can allow extra time for decisions involving conflicting information, use clear confirmation steps and watch for changes in response speed. Any operational changes would still require testing under realistic work conditions before researchers could judge their value.

Stroop tasks are useful in a pressure chamber because they need little space and can run on a compact display. Large brain scanners cannot operate easily in such environments, while repeated number tests could help researchers follow cognitive changes during different stages of compression and decompression.

Portable **EEG and event-related potential tests** could provide the next layer of evidence. EEG records electrical activity at the scalp, while event-related measurements track the brain's response to a specific signal. Combined with reaction times and accuracy, such recordings may reveal whether pressure slows early visual processing or the later stage where a person selects an answer.

Deep saturation diving supports submarine rescue, salvage work and construction below the reach of ordinary scuba operations. The Japanese chamber study shows that expert training and heliox cannot guarantee unchanged performance at 45 ATA. Measuring subtle mental strain may help diving teams prepare for the human limits that appear hundreds of meters below the surface.