Inert Gas Narcosis: Prevention

In its simplest terms, prevention comes by avoiding exposure to partial pressures of inert gas known to produce intoxication. In practice, safe diving on compressed air requires an awareness of the condition and its effect on performance and judgement at depths greater than 30 metres. The maximum depth limit for an air dive should be between 30 and 50 metres, depending on the diver’s experience and the task to be performed. Safe diving at a greater depth requires the substitution of a less narcotic agent to dilute the oxygen, such as helium, neon or hydrogen (one form of ‘technical’ diving).

There is a firm belief among divers that adaptation to IGN can develop over repeated daily exposures and that one can therefore ‘work up’ to deep dives. Several studies have shown that, although subjective adaptation can occur, measurement of standing steadiness or reaction time showed no improvement with repeated exposure3. As with alcohol, confidence is not matched by performance, thus possibly compromising safety.

Saturation at depths between 30 and 40 metres is said to allow the development of adaptation. Excursion dives to greater depths can then be made with more safety and improved work performance. A conventional working dive to 100 metres would be inconceivable using air as the breathing medium. However, operational dives may be performed to 100 metres if the excursion is from a saturated depth of 40 metres. At that depth, the diver becomes acclimatized to the nitrogen narcosis, with a progressive improvement of job performance, approaching ‘surface’ efficiency.

For most contemporary deep diving, the effect of IGN is avoided by substituting helium, or helium-nitrogen, as the diluent gases for oxygen. Oxygen cannot, of course, be used alone because of its toxicity at high pressure (see Chapter 17), but it can partially replace nitrogen in various nitrox mixtures. Hydrogen is also being used as a substitute for nitrogen and would be ideal except for the formation of an explosive mixture with oxygen.

Evidence that helium also has some narcotic effect arises from the observation that HPNS is not the same under hydrostatic pressure as it is under helium pressure. It has been postulated that both helium and oxygen need to be considered when calculating the narcotic effects of respired gases under great pressure.

Although amphetamines ameliorate narcotic slowing of reaction time, the use of drugs to reduce narcosis has, as yet, no place in diving. Conversely, divers should be warned of the risks of taking central nervous system depressant drugs, which, in the diver, may include alcohol, sedating antihistamines (in cold and sinus preparations) and anti–motion sickness drugs. These drugs may act synergistically with nitrogen in impairing performance and judgment, although this has been clearly shown only with alcohol.

Inert Gas Narcosis: Aetiology

IGN is thought to be produced by the same mechanism as general anaesthesia with gases or volatile liquids. The inert gases and most volatile anaesthetic agents are simple molecules with no common structural feature, and they do not undergo chemical change in the body to exert their effect. This property suggests that a physical rather than chemical effect must be involved, and most research is based on the hypothesis that the mechanism is the same for all agents (the unitary hypothesis of narcosis). For this reason, the considerable efforts to understand the mechanism of action for anaesthetic agents are thought to have direct relevance for IGN.

At the turn of the twentieth century, Meyer and Overton noted a strong correlation between the lipid solubility of an anaesthetic agent and its narcotic potency, and this relationship has become known as the Meyer-Overton Hypothesis. Later, in 1923, Meyer and Hopf stated ‘all gaseous and volatile substances induce narcosis if they penetrate cell lipids in a definite molar concentration which is characteristic for each type of cell and is approximately the same for all narcotics’. This means that the higher the oil-water partition coefficient (relative solubility) the more potent the inert gas. The hypothesis is that inert gas molecules are dissolved in the lipid membranes of neurons and somehow interfere with cell membrane function so that the higher the proportion of an agent dissolved in lipid, the more potent the agent as a narcotic (anaesthetic). In truth, it has long been realized there is more to anaesthetic action than this because there are some discrepancies in this relationship (Figure 15.2). For example, although both neon and hydrogen have been shown to be narcotic, neon appears to be more so despite a similar oil-water partition coefficient, and argon is about twice as narcotic as nitrogen but again has a similar oil-water solubility ratio. There are also anomalies among the volatile anaesthetics agents, but, in general, the relationship is much closer than with other physical properties. Intravenous general anaesthetic agents (e.g. thiopentone and propofol) do not fit this relationship and almost certainly produce narcosis through a different mechanism.

Graphical representation of the Meyer Overton hypothesis. Note the position of nitrogen as the least potent of the charted narcotic agents.
Figure 15.2 Graphical representation of the Meyer Overton hypothesis. Note the position of nitrogen as the least potent of the charted narcotic agents.

The lipid solubility hypothesis has been extended by the critical volume concept10. Here, the consequence of the narcotic agent dissolved in the lipid membrane is proposed to cause swelling of the membrane. At a critical volume, the swollen membrane somehow produces the clinical features of anaesthesia. Thus, there is a lipid volume change that differentiates the anaesthetized from the unanaesthetized state. Other factors, in particular pressure compressibility of the lipid, also affect volume. That some narcotic effects can be reversed by application of increased hydrostatic pressure lends weight to this hypothesis. (See also Chapter 20, on the high pressure neurological syndrome [HPNS]).

Exceptions in both animal and human studies have led to a further refinement – the multi-site expansion model11. This model postulates that expansion occurs variably at more than one molecular site and that pressure does not act equally at the same sites. Thus, hydrostatic pressure effects (see Chapter 20) or narcotic effects may predominate.

The physical theories, in general, support the concept that the site of action is a hydrophobic portion of the cell, the traditional view being that this is the cell membrane. Many studies show that membranes are resistant to the effects of anaesthetics, and other sites have been sought, such as the hydrophobic regions of proteins or lipoproteins. More recent studies have suggested that the site of action is a protein, rather than a lipid, and that narcotics act by competitive binding to specific receptors, thus affecting synaptic transmission. It has even been postulated that although impairment of cognitive function is a result of the inert gas narcotic effect, impaired motor ability is a consequence of raised hydrostatic pressure per se12. Not all experimental observations are easily reconciled, and the definitive description of the mechanism of action for volatile anaesthetic agents at the molecular level remains to be elucidated. For those interested in delving deeper into this area, see Pleuvry.

The site of action is most likely at a synaptic level, and many studies have looked at inhibitory and excitatory neurotransmitters and receptors in the central nervous system. Neurotransmitters studied include noradrenaline (norepinephrine), serotonin, dopamine, gamma-aminobutyric acid (GABA) and glycine. GABA is the most important inhibitory transmitter in the brain. Potentiation of inhibitory pathway synapse receptors (GABA receptors) is suspected to be a major component of IGN/anaesthesia, although action at a wide variety of neuronal sites is likely.

Exposure to narcosis raises extracellular dopamine in the area of the brain controlling the extrapyramidal system, at least in rats. This action may account for some of the neuromuscular disturbances of IGN. In contrast, dopamine is increased when HPNS is exhibited (see Chapter 20).

Inert Gas Narcosis: Measurement of Central Nervous System Effects

Although suitable and reliable indices of IGN are not yet available, the search continues. Such tests would be useful in predicting individual susceptibility (diver selection), comparing the relative narcotic potencies of different respiratory diluents for oxygen, delineating the role of factors other than inert gas in producing depth intoxication and monitoring the degree of impairment during practical tasks.

Attempts to quantify the effects of IGN can be roughly divided into two methods. The first is a psychological behavioural approach measuring performance on tasks such as mental arithmetic, memory, reaction time and manual dexterity. The second relies on observing a change in some neurophysiological parameter. Some representative studies are discussed to illustrate points.

Behavioural approach

The aspects of behaviour usually studied may be divided into three categories: cognitive ability, reaction time and dexterity. The cognitive functions are the most affected and dexterity the least. One early study measured the performances of 46 men on simple arithmetic tests; reaction time and letter cancellation were measured at pressures from 3.7 to 10 ATA. This study demonstrated quantitatively the previously observed qualitative progressive deterioration with increasing pressure of compressed air. It also showed that individuals of high intelligence were less affected. The impairment noted on arrival at the target pressure was exacerbated by rapid compression.

Another study using simple arithmetic tests of manual skill showed that narcosis was maximal within 2 minutes of reaching depth, and continued exposure did not result in further deterioration, but rather there was a suggestion of acclimatization. Muscular skill was much less affected than intellectual performance. Other studies involving reaction time, conceptual reasoning, memory and psychometric tests all showed progressive deterioration with increasing pressure. Narcosis has also been measured by tests of intelligence and practical neuromuscular performance.

Some work on open water divers suggested a greater impairment of performance on manual tasks at depth when anxiety was present. Plasma cortisol and urinary adrenaline/noradrenaline (epinephrine/norepinephrine) excretion ratios were used to confirm the presence of anxiety noted subjectively. Divers were tested at 3 and 30 metres at a shore base and in the open sea. Intellectual functions, as assessed by memory test, sentence comprehension and simple arithmetic, showed evidence of narcosis in both 30-metre dives, but the decrement was greater in the ocean dives, possibly because of the greater psychological stress in the open sea. Unsurprisingly perhaps, these effects have not been reproduced in the laboratory.

Many experimental protocols have been criticized because the effects of motivation, experience and learning, for example, are difficult to control. Caisson workers participating in a card-sorting test showed some impairment at 2 to 3 ATA, especially those who had relatively little exposure to pressure. However, with repeated testing, i.e. practice, this difference disappeared, and no loss of performance was noted even deeper than 3 ATA. These experiments were repeated using 80 naval subjects at 2 and 4 ATA while breathing air and helium-oxygen mixtures. The only significant impairment was found at 4 ATA breathing air.

The effects of IGN on behaviour, as measured by the psychologist, were well reviewed by Fowler, Ackles and Porlier7, and there has been relatively little work in this area since that time. More attention has been paid to the molecular mechanisms involved. Psychological studies suggest that the behavioural effects of all inert gases that produce narcosis are identical. Human performance under narcosis is explained using the ‘slowed processing’ model. Slowing is said to result from decreased activation or arousal in the central nervous system, manifested by an increase in reaction time, perhaps with a fall in accuracy. Increases in arousal, such as by exercise or amphetamines, may explain improved performance. Manual dexterity is less affected than cognitive functioning because dexterity requires fewer mental operations and there is less room for cumulative slowing of mental operations (processes). Although memory loss and impaired hearing are features of narcosis, these effects are more difficult to explain using the slowed processing model. A similar alteration in the processing of emotional experience has more recently been proposed by Löftdal and colleagues.

Studies of the subjective symptoms of narcosis have indicated that the diver can identify these symptoms and that they could relate the effect to the ‘dose’. Euphoria, as described by terms such as ‘carefree’ and ‘cheerful’, is only one of these symptoms and may not always be present. Other descriptive symptoms such as ‘fuzzy’, ‘hazy’ (state of consciousness) and ‘less efficient’ (work capability) and ‘less cautious or self-controlled’ (inhibitory state) may be more reliable indicators of effect on performance.

Behavioural studies have cast doubt on some traditional concepts of narcosis. True adaptation to narcosis has not been found in many performance tests. Where adaptation has been found, it is difficult to distinguish learned responses or an adaptation to the subjective symptoms from physiological tolerance. Carbon dioxide probably has additive and not synergistic effects in combination with nitrogen and probably acts by a different mechanism. Behavioural studies have not been able, so far, to demonstrate clearly the potentiating effects on IGN of anxiety, cold, fatigue, anti–motion sickness drugs and other sedatives (except alcohol).

Neurophysiological changes

Attempts have been made to confirm the subjective experiences and obtain objective evidence of performance decrement, with some neurophysiological parameter. The investigations included electroencephalographic records of subjects exposed to compressed air in chambers. Contrary to the expected findings of depression, features suggesting cortical neuronal hyperexcitability were noted at first. These included an increase in the voltage of the basal rhythm and the frequent appearance of low-voltage ‘spikes’ elicited by stimuli that do not have this effect at 1 ATA. Experiments in which the partial pressures of oxygen and nitrogen were controlled showed that in compressed air these changes are caused by the high oxygen partial pressure. If nitrogen-oxygen mixtures containing 0.2 ATA oxygen are breathed, these changes are absent. The depressant effects of nitrogen are then revealed. These consist of a decrease in the voltages of the basal rhythm and the appearance of low-voltage theta waves.

Blocking of electroencephalographic alpha rhythm by mental activity can be observed in half of the population. The observation that there is an abolition of this blocking on exposure to pressure introduced the concept of ‘nitrogen threshold’. It was found that the time to abolition of blocking was inversely proportional to the square of the absolute pressure (T is proportional to 1/P2) for an individual, although there was marked variation among subjects. In some persons abolition of blocking was noted at depths as shallow as 2.5 ATA, where no subjective narcosis was evident.

Flicker fusion frequency was investigated in an attempt to obtain a measurement that could be applied to the whole population. Subjects were asked to indicate when the flickering of a neon light, at a steadily increasing rate, appeared continuous. This is termed a ‘critical frequency’ of flicker. After a certain time at pressure, the critical frequency dropped. The same relationship, T is proportional to 1/P2, resulted. Critical flicker fusion tests have been adapted for the in-water environment and continue to be used in the context of IGN assessment.

A more direct measure of central nervous system functioning may be obtained by observing the effect of inert gas exposure on cortical evoked potentials. Evoked potentials are the electroencephalographic response to sensory stimuli. A depression of auditory evoked responses on exposure to hyperbaric air has been shown to correlate with the decrement in mental arithmetic performance under the same conditions. The conclusion was that auditory evoked response depression was an experimental measure of nitrogen narcosis. However, other work was unable to support this hypothesis and concluded that there is a complex relationship among hyperbaric oxygen, nitrogen narcosis and evoked responses.

Auditory evoked responses as a measure of narcosis are problematic because of sound alteration with pressure and the ambient noise during hyperbaric exposure. Visual evoked responses (VERs) have been used in an attempt to produce more reliable information. VERs were studied in US Navy divers, and reliable and significant differences were reported while the divers were breathing compressed air versus helium-oxygen mixtures at pressure. A further study using VERs during a shallow 2-week saturation exposure with excursion dives suggested that some adaptation to narcosis occurred, but it was not complete. Reduction of frequency and amplitude of alpha activity when compared with pre-exposure and post-exposure surface levels were also noted. Nevertheless, the value of current methods of measurement of IGN, by the use of neurophysiological changes, is questionable.

Inert Gas Narcosis: Clinical Manifestations

Martini’s law illustrated. Diving on air at 60 metres is a dangerous challenge.

Martini’s law: Each 15-metre (50-foot) depth is equivalent to the intoxication of one martini.

Although there is marked individual variation in susceptibility to IGN, all divers breathing compressed air are significantly affected at a depth of 60 to 70 metres (Figure 15.1). The minimum pressure producing signs is difficult to define, but some divers are affected subjectively at less than 30 metres.

Martini’s law illustrated. Diving on air at 60 metres is a dangerous challenge.
Figure 15.1 Martini’s law illustrated. Diving on air at 60 metres is a dangerous challenge. (Illustration courtesy of

The higher functions, such as reasoning, judgement, recent memory, learning, concentration and attention are affected first. The diver may experience a feeling of well-being and stimulation similar to the overconfidence of mild alcoholic intoxication. Occasionally, the opposite reaction, terror, develops. This is more probable in the novice who is apprehensive in this new environment. Further elevation of the partial pressure of the inert gas results in impairment of manual dexterity and progressive deterioration in mental performance, automatisms, idea fixation, hallucinations and, finally, stupor and coma. Some divers complain of a restriction of peripheral visual field at depth (tunnel vision). They are less aware of potentially significant dangers outside their prescribed tasks (perceptual narrowing). More recently, abnormal emotional processing has been described, with a suggestion that the emotional responses to threat are muted with increasing IGN.

From a practical point of view, the diver may be able to focus attention on a particular task, but the memory of what was observed or performed while at depth may be lost when reporting at the surface. Alternatively, the diver may have to abort the dive because of failure to remember instructions. Repetition and drills can help overcome these problems through a ‘practice effect’. Conversely, anxiety, cold, fatigue, sedatives, alcohol and other central nervous system depressant drugs aggravate narcosis.

Nitrogen narcosis has often been likened to alcoholic intoxication, especially the euphoria, lightheadedness and motor incoordination. There is some evidence that correlates subjective feelings of alcohol consumption and IGN, especially the variation in intensity experienced among individuals. In one elegant experiment reported in 2008, Hobbs2 found evidence that heavier drinkers did not show a reduction in the effects of nitrogen narcosis at depth, but they did show tolerance for the combined effects of narcosis and alcohol. At much greater depths the parallel with general anaesthetic agents is probably closer.

Some of the reported observations at various depths breathing compressed air are shown in Table 15.1.

Some observations on the effects of exposure to compressed air at increasing pressure/depth

The narcosis is rapidly evident on reaching the given depth (partial pressure) and is not progressive with time. It is said to be more pronounced initially with rapid compression (descent). The effect is rapidly reversible on reduction of the ambient pressure (ascent).

Other factors have been observed to affect the degree of narcosis. Cold, reduced sensory input, and both oxygen and carbon dioxide disturbances are interrelated in impairing the diver’s underwater ability. In experimental conditions, with an attempt to control variables, alcohol and hard work have been shown to enhance narcosis. Moderate exercise and amphetamines may, in certain situations, reduce narcosis, but some studies have conversely suggested unpredictable or increased narcotic effects with amphetamines. Increased carbon dioxide and nitrogen tensions appear to be additive in reducing performance. Task learning and positive motivation can improve performance. Frequent or prolonged exposure produces some acclimatization, but this may reflect a reduction in psychological stress rather than representing true adaptation. For example, Hamilton and associates3 have experimental evidence to suggest that the reported experience of adaption is more subjective than behavioural.

Direct pathological injury to the diver as a result of the high pressure of inert gas is unlikely. The danger is rather a result of how the diver may react in the environment while under the narcotic influence of nitrogen. Impaired judgement can lead to an ‘out-of-air’ drowning sequence, with no other apparent cause of death found. The diver affected by IGN may also be at increased risk of insidious hypothermia (see Chapter 28) because of decreased perception of cold and decreased shivering thermogenesis4,5. Jacques Cousteau6 shared his experience of nitrogen narcosis (Case Report 15.2).


We continue to be puzzled with the rapture of the depths, and felt that we were challenged to go deeper. Didi’s deep dive in 1943 of 210 feet had made us aware of the problem, and the Group had assembled detailed reports on its deep dives. But we had only a literary knowledge of the full effects of I’ivresse des grandes profondeurs, as it must strike lower down. In the summer of 1947 we set out to make a series of deeper penetrations.

…I was in good physical condition for the trial, trained fine by an active spring in the set, and responsive ears. I entered the water holding the scrap iron in my left hand. I went down with great rapidity, with my right arm crooked around the shotline. I was oppressively conscious of the diesel generator rumble of the idle Elie Monnier as I wedged my head into mounting pressure. It was high noon in July, but the light soon faded. I dropped through the twilight, alone with the white rope, which stretched before me in a monotonous perspective of blank white signposts.

At 200 feet I tasted the metallic flavour of compressed nitrogen, and was instantaneously and severely struck with rapture. I closed my hand on the rope and stopped. My mind was jammed with conceited thoughts and antic joy. I struggled to fix my brain on reality, to attempt to name the colour of the sea around me. A contest took place between navy blue, aquamarine and Prussian blue. The debate would not resolve. The sole fact I could grasp was that there was no roof and no floor in the blue room. The distant purr of the diesel invaded my mind – it swelled to a giant beat, the rhythm of the world’s heart.

I took the pencil and wrote on a board, ‘Nitrogen has a dirty taste’. I had little impression of holding the pencil, childhood nightmares overruled my mind. I was ill in bed, terrorised with the realisation that everything in the world was thick. My fingers were sausages. My tongue was a tennis ball. My lips swelled grotesquely on the mouth grip. The air was syrup. The water congealed around me as though I were smothered in aspic.

I hung stupidly on the rope. Standing aside was a smiling, jaunty man, my second self, perfectly self-contained, grinning sardonically at the wretched diver. As the seconds passed the jaunty man installed himself in my command and ordered that I unloose the rope and go on down.

I sank slowly through a period of intense visions.

Around the 264 foot board the water was suffused with an unearthly glow. I was passing from night to an imitation of dawn. What I saw as sunrise was light reflected from the floor, which had passed unimpeded through the dark transport strata above. I saw below me the weight at the end of the shotline, hanging twenty feet from the floor. I stopped at the penultimate board and looked down at the last board, five metres away, and marshalled all my resources to evaluate the situation without deluding myself. Then I went to the last board, 297 feet down.*

The floor was gloomy and barren, save for morbid shells and sea urchins. I was sufficiently in control to remember that in this pressure, ten times that of the surface, any untoward physical effort was extremely dangerous. I filled my lungs slowly and signed the board. I could not write what it felt like fifty fathoms down.

I was the deepest independent diver. In my bisected brain the satisfaction was balanced by satirical self-contempt.

I dropped the scrap iron and bounded like a coiled spring, clearing two boards in the first flight. There, at 264 feet, the rapture vanished suddenly, inexplicably and entirely. I was light and sharp, one man again, enjoying the lighter air expanding in my lungs. I rose through the twilight zone at high speed and saw the surface pattern in a blaze of platinum bubbles and dancing prisms. It was impossible not to think of flying to heaven (Cousteau JY. The Silent World. London: Reprint Society; 1954).

* 297 feet is 90.5 metres (don’t try repeating this experiment!).

Inert Gas Narcosis: History

Inert gas narcosis (IGN) was comprehensively reviewed in 2003, and interested readers are referred to that work for more detail1. The first recorded description of symptoms suggestive of air intoxication related to hyperbaric exposure was by Junod, who, in 1835, reported that ‘thoughts have a peculiar charm and in some persons, symptoms of intoxication are present’. He was conducting research into the physiological effects of compression and rarefaction of air. J. B. Green in 1861 observed sleepiness, impaired judgement and hallucinations in divers breathing compressed air at 5.8 ATA, sufficient to warrant an immediate return to the surface. Paul Bert, in 1878, also noted that divers became intoxicated at great depth. In 1903, Hill and McLeod described impairment of intellectual functioning in caisson workers at 5.5 ATA. Damant, in 1930, likened the mental abnormalities and memory defects observed in men at 10 ATA to alcoholic intoxication and postulated that it was caused by the high partial pressure of oxygen.

Hill and Phillips suggested in 1932 that the effects could be psychological as a result of claustrophobia or perhaps caused by impurities in the air from the compressors (Case report 15.1). The Royal Navy appointed a committee to investigate the problems of deep diving and submarine escape, and their report in 1933 contained a section entitled ‘semi-loss of consciousness’. Between 7 and 11.6 ATA, divers answered hand signals but in many cases failed to obey them. After return to surface, the divers could not remember the events of the dive. It was noted that all divers regained full consciousness during the return to 1 ATA. The report also noted great individual variation in divers’ reactions but was unable to elucidate the problem.


‘You notice the dark more although it may not be darker; the light is a comfort and company. You notice things more if there is nothing to do; I get comfort from seeing the fish, it takes your mind off everything else’.

When asked for a description, an old hand at diving gave the following account: ‘You have to be more careful in deep water; in deep water you know that you are concentrating… You think of each heave as you turn a spanner… If you go down with a set purpose it becomes an obsession; it will become the main thing and you will forget everything else’. He described how he thinks very deliberately; he says, ‘I have finished my job, what shall I do next? – Of course, I have finished and now I must go up’. He described how he was aware of every action: ‘If my hand goes out I think of my hand going out’. He gave the following as an analogy: ‘if I saw a thing of value, say half-a-crown, in the street, I would pick it up. Down below I would look at it and think, “What is that, shall I pick it up? Yes, I will pick it up” and then I feel my hand go out’.

‘I left the ladder determined to get to the bottom. At 250 feet I got a recurrence of the tingling and a feeling of lying on my back. I decided to rest a couple of minutes and then go on. I slid 10 feet and felt I was going unconscious. I made signals to be pulled up and kept repeating them. I lost the use of my limbs and let go everything. While hanging on to the rope I saw my own face in the front glass; it was outside the glass and looked all greenish; I was dressed in my shore-going suit. I heard the order, “Pull the diver up”, again and again, as if someone in the suit was saying it’.

‘Suddenly I came over rather “funny”; it was a distinct “different” feeling; I stood up, the tank wire in my right hand, and thinking it was a touch of CO2, I began to breathe deep and hearty, thinking of course that in a couple of minutes I would be able to resume work. Then I seemed to go quite limp, a feeling of “no life or energy”. This was new and strange to me, whether it was a part of CO2, I didn’t know, because I had never experienced a real dose of CO2; anyhow, after stopping and doing the drill for CO2, I thought I would be alright, but suddenly something definitely seemed to say – snap inside my head and I started to, what I thought, go mad at things’.

Description of interview of above diver after an aborted deep dive:

Practically no hypnoidal effort was required to produce the horrors of that morning’s dive, and the picture of stark, mad terror….left an impression which is very difficult to describe. The impression was of sitting in the stalls and watching the acting of Grand Guignol. To such a pitch did he arouse his emotions that he clawed his face to remove the imaginary face-glass and tore his clothes which he mistook for his diving suit (Hill I, Phillips AE. Deep-sea diving. Journal of the Royal Navy Medical Service 1932;18:157–173).

It was not until 1935 that Behnke, Thomson and Motley proposed the now generally held theory of the cause of this compressed air intoxication. They stated that the narcosis was the result of the raised partial pressure of the metabolically inactive gas, nitrogen. At a depth of 30 metres (4 ATA), compressed air produced a state of ‘euphoria, retardation of the higher mental processes and impaired neuromuscular co-ordination’. This effect was progressive with increasing pressure so that at 10 ATA, stupefaction resulted. Unconsciousness developed between 10 and 15 ATA. They also invoked the Meyer-Overton hypothesis (see the later section on aetiology) to relate the narcotic effect to the high ratio of solubility of nitrogen in oil to water. It was not long after this major breakthrough that Behnke and Yarbrough reported that the substitution of helium for the nitrogen in compressed air eliminated narcosis.

The nitrogen partial pressure theory was not universally accepted. The 1933 Deep Diving Committee Report had raised the possibility that carbon dioxide retention was implicated. Case and Haldane, in 1941, reported that the addition of carbon dioxide to compressed air worsened the mental symptoms, although up to 6 per cent concentrations at 1 ATA had little mental effect. Bean, in 1947, demonstrated a reduction in arterial pH during compression and later also showed increased alveolar carbon dioxide concentrations. He explained these changes as being caused by reduced diffusion of carbon dioxide in the increased density of the air. He postulated that carbon dioxide was an alternative cause of depth narcosis. Seusing and Drube later supported Bean’s views as recently as 1961. Also in 1961, Buhlmann believed that increased airway resistance led to hypoventilation and hypercapnia.

Rashbass in 1955 and Cabarrou in 1959 had already refuted the carbon dioxide theory, by observing signs of narcosis despite methods to ensure normal alveolar carbon dioxide levels. Later work (by Hesser, Adolfson and Fagraeus) showed that the effects of nitrogen and carbon dioxide are additive in impairing performance. Normal arterial carbon dioxide and oxygen levels, while the diver is breathing air and helium-oxygen at various depths, demonstrate the key role of nitrogen in the production of this disorder and the relative insignificance of carbon dioxide.

Inert Gas Narcosis: Introduction

Inert gas narcosis refers to a clinical syndrome characterized by impairment of intellectual and neuromuscular performance and changes in mood and behaviour. It is produced by an increased partial pressure of some inert gasses. In compressed air exposure, these changes, which have been observed for more than 100 years, are caused by nitrogen. The effects are progressive with increasing depth but not with increasing time at the same depth. The word ‘inert’ indicates that these gases exert their effect without undergoing metabolic change in the body, rather than inert gas in the biophysical sense.

Similar effects have been described with other metabolically inactive gases such as the rare gases (neon, argon, krypton, xenon), hydrogen and the anaesthetic gases, although at different partial pressures. Xenon is ‘anaesthetic’ at sea level and is used in some parts of the world as an anaesthetic agent. No narcotic effect of helium has been directly demonstrated at currently attainable pressures, although some narcotic properties have been postulated.

The ‘inert’ gas in compressed air is nitrogen, and its effects are also called nitrogen narcosis, depth intoxication, ‘narks’ and rapture of the deep (I’ivresse des grandes profondeurs), the term coined by Cousteau. The narcosis, although highly variable, places a depth limit to safe diving with compressed air at approximately 40 to 50 metres. Effective work at greater depth requires the substitution of a less narcotic respiratory diluent such as helium or hydrogen.

Many ‘unexplained’ scuba deaths may have been associated with nitrogen narcosis.