Heating and humidification in the normal respiratory tract

This chapter is most relevant to Section F10(iv) from the 2017 CICM Primary Syllabus, which expects the exam candidates to be able to "define humidity and give an outline of the importance of humidification." Humidity is defined in the chapter on the physics of humidity and evaporation, and supplemental artificial humidification s described in the chapter on active circuit humidification equipment. This page deals specifically with the gas heating and humidification functions of the human respiratory tract in its native unmolested state. 

The topic of humidification had appeared several times in the Part One exam. 

• Question 3 from the second paper of 2012
• Question 13 from the second paper of 2010
• Question 7(p.2) from the second paper of 2009

These have been questions dealing predominantly with humidification, and though heating is not explicitly mentioned in the question, the examiners clearly expected it ("Whilst the question asked ‘humidity’ and not temperature, correct definitions in the first part would have dictated a joint outline of both.") 

In summary:

The best peer-reviewed article for this topic is actually the ancient paper by Walker & Wells (1961) but unfortunately, it is paywalled. Something more recent but no less paywalled is Jackson (1996). Davis' Basic Physics and Measurement in Anaesthesia has an excellent section on heat loss and moisture loss by vapourisation (around p.132 of the 4th edition). For the freegan, the information exists in numerous distracted resources around the web; no specific single source is good enough as a stand-alone reference, but Boys & Howells (1972) come closest as they cover pretty much everything required to answer the CICM Part One SAQs.

Temperature and humidity of ambient hospital air

Starting from the premise that ICU trainees are unlikely to be managing patients in the tundra as a part of their routine practice (retrieval rotations notwithstanding), this chapter will begin with a discussion of the characteristics of standard hospital-grade room air.

This is one of those scenarios where, for parameters like heat and temperature, it is as difficult to determine what they should be as it is to establish what they routinely are. In 1977, Smith & Rae published on the "Thermal comfort of patients in hospital ward areas", suggesting that they prefer 21.5-22° C and a range of 30%-70% relative humidity. It is difficult to know whether this preference is acknowledged by hospital facility design.   A bit of Googling reveals a local (NSW Health) Engineering Services Guideline (2016) which gives these recommended ranges for temperatures and humidities in NSW hospitals:

Relative Humidity and Temperature Guidelines

Room	Relative humidity	Designed temperature range
Ward area	no specific range	21-24° C
Operating theatre	35-60%	16-27° C
Intensive care	35-60%	21-24° C
Neonatal ICU	35-60%	22-26° C
Burns unit	35-95%	21-32° C
Random hospital corridor	no specific range	no specific range

In short, we can see that at the lowermost point, hospital designs factor in a temperature of 16° C and a relative humidity of 35% (in the coldest operating theatre), which corresponds to an absolute humidity of only 5g/m³. At the highest part of the range, in the most tropical part of the burns unit, the absolute water content of air would be 32g/m³ (95% humidity, 32° C). Assuming that most ICU areas will hover at some midpoint of the temperature and humidity range, one can calculate that the absolute content of the air our patients normally inhale is 10g/m³ (50% humidity at 22° C). All of these numbers, by the way, are generated using the excellent Planetcalc humidity calculator, into which all sorts of numbers can be plugged. For instance, at 150° C and relative humidity of 100% the absolute humidity of superheated steam is 2.6kg/m³.  Presumably, that's the sort of atmosphere-of-Venus-like environment hospital designers deign suitable for foyers corridors and thoroughfares because no specific climate guidelines exist for those shared areas.

Heating and humidification of inspired gas

"It is of interest that overwhelming hyperventilation with a dry atmosphere had no effect upon the climatic conditions of the alveolar cavity", remarked Raynald Déry (1971) after creating bronchoalveolar fistulas in anaesthetised dogs and measuring the humidity of the escaping gas. Dogs ventilated with a wide variety of gas mixtures at different temperatures were still able to heat and humidify their respiratory gas mixture well enough so that alveolar gas humidity remained constant, heated to body temperature. On the way towards the alveolus, the gas became heated by the airway structures to a point where it became the same temperature as the rest of the body (i.e. it became isothermic with the core body temperature). At the same time, the relative humidity increased to 100%.  The point in the respiratory tract at which the inspired gas mixture achieves this alveoli-like temperature and humidity is usually called the isothermic saturation boundary, described in this classic diagram stolen shamelessly from Déry (1971).

In summary, within a massive range of conditions, the temperature and relative humidity of the alveolar gas mixture increases to  37° C and 100%, corresponding to about 47g/m³ total water content, and this occurs somewhere just past the carina (5cm distal to it, or in 2nd-4th generation bronchi according to the variable position given by different textbooks). Heating and humidification are somewhat separate processes; the mucosa, as it releases water vapour, donates heat to the process of evaporation and is cooled, but the heat is trapped as latent heat of vapourisation which does not change the temperature of the air. It is released again by condensation which occurs when expired air flows over the cooled mucosa.  

Reclamation of heat and moisture during expiration

After all the effort and energy expended on warming and wetting the inspired gas mixture, it makes some sense that the upper respiratory tract should try to reclaim some of the water and heat. It is less efficient at this dehumidification than it is at humidification.  Cole (1954) inserted a thermister up the nose of an undescribed "subject" and described the temperature of the escaping breaths. In a variety of ambient conditions, the expired air had a relatively stable temperature of around 32° C degrees. This exchange occurs because the air, coming in through the airways, cools their mucosal surface; on the way out the air is warmer than the surrounding mucosa, and donates some of its heat back to the walls of the airways.

So, assuming we start from 47g/m³ at 37° C, if expired gas ends up at 100% relative humidity its water content will be 34g/m³, with the remaining 13g/m³ precipitating as a sort of dew on to the respiratory mucosa. Some heat and moisture are reclaimed thereby. The rest is lost from the organism, which seems wasteful but is in fact a minimal contributor to the total body heat loss.  Davis' Basic Physics and Measurement in Anaesthesia (p. 127) calculates that with a minute volume of 7L/min, the total heat loss is 118 J/min, or around 2W - compared to the 80W of basal heat loss.

The recondensation of heat and water is probably more important for species which are constantly exposed to perverse conditions. For example, Schmidt-Nielsen et al (1970) report that some animals like the kangaroo rat (which live in water-poor environments) can reclaim 70-80% of their exhaled moisture. In these animals, reclaimed heat amounts constitute a savings of 16.1 % of the total metabolic heat production.

Influence of ambient temperature

The respiratory tract is extremely efficient at heating and humidifying inspired gas, even in the face of very low ambient temperatures.  For an extreme example, Moritz et al (1945) ventilated dogs with a vacuum-jacketed Dewar cannula which delivered superchilled oxygen mixture at -100° C. The dogs' lower tracheal temperatures did not drop any lower than 18° C.  "In all of this group of animals the air leaving the lungs was within 1 or 2 degrees of normal", the authors noted. Of the dogs who were not sacrificed for histology, the rest were "lively, happy and not significantly disturbed by the experimental procedure" even after breathing the superchilled gas or up to 133 minutes.

Of course, for obvious reasons, the Australian Intensive Care trainee will probably be more interested in the other temperature extreme. For this, we also have some data, again from earlier in the 20th century.  Cole (1954) for example wrote about the temperature of exhaled air in Turkish sauna environments (40° C), noting that it was roughly the same as core body temperature (i.e. the body cooled the inhaled gas on the way through the respiratory tract). The hot air on the way into the lungs would therefore have warmed the respiratory mucosa, preventing the normal expiratory condensation of dew from taking place.

The reclamation of moisture, therefore, suffers in hot conditions. Even though the diagram below was borrowed from Schmidt-Nielsen et al (1970) who were discussing the cactus wren, it can still be used to illustrate the change in the amount of water recovered from respiratory gases as the ambient heat increases. In short, it is halved between 15 and 30 degrees. This makes sense: the hotter it gets the more moisture you lose.

Alterations in heat and moisture exchange in exercise and disease

The mechanisms responsible for humidification of inspired gas are made all the more remarkable by their extreme resilience against any variation in respiratory function.  Déry hyperventilated some of his dogs to over 30L/min minute volumes, using cold dry gas straight from the tank; the alveolar humidity and temperature did not budge. Instead, the isothermic boundary (normally about 5cm from the carina) just moved deeper into the lung.

Theoretically, the humidification functions of the conductive airways continue throughout the generations of bronchi all the way up to the respiratory bronchioles. As ventilation increases or the temperature of inspired gas decreases, the isothermic boundary moves further and further into the lung. According to the 8th edition of Nunn's, at minute volumes of over 50L/min the isothermic boundary has moved out of first-generation bronchi and into airways of 1 mm diameter. This factoid is repeated extensively throughout the literature (eg. it also appears verbatim in Webster & Galley's Anaesthesia Science) but it is difficult to determine where it came from, who was ventilated with 50L/min and how the humidity was measured in those tiny little bronchi.

According to Walker & Wells (1961), under normal conditions, an adult loses 250 ml of water and 350 kcal of heat in their expired air per day. Under less normal conditions (eg. low humidity, increased respiratory rate and tachycardia)  water losses can increase. The influences of these things on the humidity of exhaled air and the water losses by respiration are explored to some considerable depth in a 2012 article by Zieliński et al. The writing sparkles, but is Polish. In the translated abstract, the English-speaking reader can find interesting data, particularly regarding the effect of temperature and ambient humidity on water loss. 

For instance, the authors note that at 35°C and 75% humidity, hourly respiratory water loss is approximately 7ml; but when you lower the temperature to -10°C, the humidity drops to 25% and water loss increase to 20ml/hr. Furthemore, heart rate (i.e. cardiac output) influences the rate of water loss enormously by increasing the delivery of water to the gas exchange surfaces, and a tachycardic person with a heart rate of 140bmp may actually lose up to 60-70ml/hr.  This has obvious implications for situations where a person may be compelled to exercise for prolonged periods  (eg. some sort of sadistic charity fun run). Run a six hour marathon and you may have lost up to half a liter of water purely by breathing.

Structural features of the airways which facilitate humidification

To work this magic, we need specific structures which have, as their main goal, the increase in air turbulence. Turbulent flow allows more of the inhaled air to come into contact with the respiratory tract mucosa, and allows mixing of the air which promotes heat exchange. This "turbulent convection" is essential because air is a poor conductor of heat, which means we cannot rely on conduction or radiation.

This need to increase turbulent flow is reflected in the characteristics of nasal turbinate design among the mammals.

The diagram above is modified from Hillenius (1992). As one can extrapolate from the crossections, one can expect the northern elephan seal to be an absolute champion of airway gas humidification, and indeed this appears to be the case. They hold the record for the largest nasal mucosal area (in excess of 3000 cm²) and are able to reclaim over 80% of the exhaled moisture, with the expired gas achieving a temperature very close to ambient by the time it reaches the nares.

Effects of intubation and oxygen therapy on humidification

In the ICU, it is often necessary to bypass the upper airway with some sort of tube, be it endotracheal or tracheostomy.  The normal humidification mechanisms are therefore bypassed. This has the following ill efects:

• Increased heat loss by respiration
• Increased water loss by humidification
• Impaired humidification, if no special effort is made to heat and humidify the inspired gas mixture

Beyond these factors, there are aso the effects of using a respiratory gas mixture which is even less humid than room air, which comes from the hospital gas supply system.  In hospital, a person is exposed to a variety of perverse gases, which tend to come out of wall-mounted spigots. The spigots are themselves portals to a nighmarish labyrinth of pipes, all terminating in the bowels of the hospital where a huge pressurised tankers of liquid air and oxygen are carefully contained. The liquid O₂,  having just undergone a phase change, is completely devoid of water: it has zero humidity. At room temperature, it is therefore very thirsty.

Now, imagine having your lungs irrigated with this substance. The volume of gas exiting the respiratory circuit is enormous - it may be 5L per minute at rest, or more - up to 15L/min - in somebody exercising, or in a critically patient who is fighting for breath. Therefore around 300L of gas washes out of the lung every hour, or 7200L in a day. The   respiratory system can humidify gas to 80-90% by the time it reaches the carina, and provided one is breathing with their nose is is possible to reclaim some of that water. Bypassing the nose with an endotracheal tube maximises water loss. At 37°C, the water content of air is around 44mg/L, or 0.044ml.  It is therefore possible to calculate the water losses of an intubated patient who is for some reason exposed to some sort of inhuman "dehumidified" circuit, where perfectly dry gas was inhaled and perfectly saturated humid gas was exhaled. Loss of water would total 0.22ml every minute, or about 13.2ml per hour, or about 316.8ml every day.

Paradoxically, only relatively healthy patients are gassed with bone-dry wall oxygen; it tends to go directly up their nose, and it causes mayhem up there. However, one does not need to be breathing perfectly dry "wall gas" to develop complications of dehydration. Cilial paralysis and reduced rates of mucus flow occur below 40%  humidity.  We know this directly from cruel dog experiments (many dog tracheas were sliced and examined under the SEM), as well as from abundant human experience. In his classic paper from 1952, Lassen complained that his polio patients were getting "tube incrustation" unless humidified circuit were used. Indeed, mucus clearance in the bronchi is greatly impaired by dry air. At 50% humidity, mucus flow essentially stops in the majority of people (insofar as they are represented by anaesthetised greyhounds).

Again, perhaps this is a digression. The bottom line is that local dehydration impairs bronchial immunity. An increased susceptibility to infection is the consequence. The Bang family ( Drs. Betsy and Frederick Bang ) published on this subject specifically (1963, Annals of the New York Academy of Sciences). Their animal of choice was the day-old chicken, presumably because there was nothing more cute, but also because other mammals are unsuited for such experiments as their nasal structures are specialised for a more refined olfaction than the crude human nose (too full of goblet cells, too complex an architecture of the nasal turbinates, etc etc). The article is complete with beautiful sketches of dead and dying goblet cells. Betsy and Frederick have actually published about 40 papers together, mostly on the pathology of airway linings,  with Betsy as the first author in the majority.