Health & Clean Air Home Health & Clean Air Newsletter Health & Clean Air Newsletter

Dr. David V. Bates,
Professor Emeritus of Medicine, University of British Columbia


My invitation to give this lecture included the suggestion that I should pick up some topic that I had been interested in for most of my professional life, and follow its development to the present. I chose to talk on ozone because I do have some milestones to chart the progress of my interest, and because, as I shall show you, it is becoming more relevant rather than less as time goes on. I shall insert some notes on the development of interest in the small airways of the lung; the reason for this parallel history will become evident as we come up to date.

The story of the “discovery” of ozone by Schonbein in Basel in the middle of the last century is an interesting one. He had worked in England teaching German at a boarding school, and at that time met Michael Faraday. Here he must have learned that an electric spark was often followed by a curious smell; when he got back to Basel he became later a professor of Chemistry and was able to show that it was ozone – O3 - that was being formed. He did a very thorough job of investigating the properties of this gas, and discovered that it was intensely irritant. Its bactericidal properties became the center of attention later in the century, and W.S,Gillbert wrote some doggerel verse calling it “Nature’s Great Cleanser”. This was unearthed by Bill Linn.

I believe it was the German V2 rocket programme that led to the recognition of the ozone layer, starting at about 30, 000 feet and rapidly increasing to a maximum at about 50,000 km before waning. I am not sure of the history of this, but I was first shown German data on the ozone layer on a visit to the Royal Aeronautical Establishment at Farnborough. Our research group, led by Prof. Ronald Christie at Bart’s Hospital was one of the few research groups with an interest in the lung, and we exchanged relevant information. It was there that I learned that the foam rubber in the passenger oxygen masks in the de Havilland “Comet” had deteriorated rapidly, and that the Dunlop company had suggested that it must be ozone which had caused this disintegration. The Comet was the first aircraft to fly routinely above 30,000 feet. We all realized that we knew nothing of the toxicity of ozone if inhaled at sea level; but this research was made difficult by the lack of any reliable instrument to measure the gas concentration.

When I went to the Royal Victoria Hospital in 1956 with a mandate from Ronald Christie to start a respiratory research unit, the problem of ozone toxicity was one of the questions I took with me. The development of the “MAST” ozone analyzer in the early nineteen sixties was a great step forward. I wanted to measure the ozone levels in the latest generation of jet aircraft, and fortunately the Chief Engineer of Air Canada at that time, Jack Dyment, was a near neighbour in Montreal. I learned that Air Canada was about to take delivery of their first DC-8, and with him I planned an experiment whereby we would put rubber bands in different places on the aircraft and then replicate the hours the aircraft had spent above 30,000 feet in a box at ground level using an ozone meter; but first I had to go to the Dominion Rubber company in Montreal and get them to supply me with rubber bands each about half an inch wide and six inches long. There had been several other reports of using rubber cracking as an indicator of ozone concentration. The rubber had to be specially prepared so that it contained no anti-oxidants.

All of this is described in my paper in the Journal of Aerospace Medicine which came out in 1962 [1]. We found the highest ozone level in the tube located in the nosewheel compartment, and the band cracking in the tube taped under the table used by the navigator was slightly less. The tube at the rear of the passenger cabin was affected, but the least of the three. We estimated that levels in the cockpit would be between about 50 parts per billion for the 270 hours the aircraft had spent over 30,000 feet. Some years later, the FAA in the United States together with Scandinavian Airlines began to make systematic observations of ozone in the aircraft on the Seattle-Copenhagen run and these broadly confirmed these calculations.

My first paper on the effects of breathing ozone came out in 1964. When I became Chairman of the Department of Physiology at McGill in 1967, I had space to build a plastic chamber and thus could study the effects of breathing ozone while exercising, which turned out to be an important innovation.[2].

Running in parallel with this work on ozone, we were developing and exploiting the use of radioactive xenon133 in studying regional lung function. These studies led to the recognition of regional factors involved in nonuniform ventilation distribution – a topic that had been of interest to me since I started in pulmonary research in 1948. Using inhaled boluses of Xe133, and studying the effects of different factors such as body position and age on gas distribution, it became clear that in normal subjects, due to the shape of the pressure-volume curve of the lung, regional airway closure was occurring in the most dependent parts, and this led us to begin to think about small airway function in general and airway closure in particular [3]. At the same time, others were getting interested in airway resistance in the small airway region of the lung, and in studies of comparative morphometry, it was clear that the respiratory bronchioles should be viewed as a target region of the lung.

I can remember being stimulated also by dosimetric calculations being modeled by Fred Miller and his colleagues [4,5], which showed beyond much doubt that the highest concentration of inhaled ozone, in terms of concentration per square centimeter of the wall, would be in the small airways or terminal bronchioles, and by 1973 I was beginning to talk about small airways as a target area for gases and particles [6].

All of this work, together with the beginning of quantitative morphometry, led me to give a talk at the symposium honouring Julius Comroe on his retirement [7], which I called “The Last Link in the Chain”. The idea behind this was that we had spent a lot of time analyzing gas exchange at the alveolar level and studying diseases in which the prime impact was on gas exchange, and we had come to understand the changes in large airways which led to hyper-responsiveness and mucus hypersecretion, but the new data emerging on small airways was the last link in the chain and constituted a new area of research of importance as a bridge between these two regions. This lecture was published in 1977. [7].

Another milestone in this journey was a major symposium held in Copenhagen in 1979 [8]. This brought together everyone who at that time was interested in some aspects of small airway pathophysiology, and firmly established that “The Last Link in the Chain” that I had described two years earlier was indeed filling an important gap in our understanding of the lung as a whole. The symposium proceedings ran to over 250 pages.

I should perhaps pause here to note that the ozone in aircraft problem had been the subject of intense enquiry as the aircrew had begun to complain of symptoms which they (mostly correctly) attributed to ozone [9]. Modifications to the compressors controlling cabin pressure had resulted in cabin ozone levels that were much lower than what I made measured.

Meanwhile, the field of ozone research had continued its active development. I took an opportunity to be a visiting scholar at the Health Effects Laboratory of the EPA at Chapel Hill with Phil Bromberg and my ex PhD student, Milan Hazucha. We planned a new study to clarify the sequence of events when ozone is inhaled. We were very slow in getting this work out as it didn’t appear until 1989, almost three years after the experimental work had been done [10]. We found that the initial effect of ozone was to limit the forced inspiration, probably via stimulation of the C-fibre network, and a spinal mediated reflex. The consequent decline in FVC was accompanied by a decline in FEV1. Thereafter a major effect was a direct one on the small airways, which was quite slow to resolve after exposure had ceased.

It is the case in all research fields that some periods are characterized by a pause in new information and discovery. The 1980’s constituted such a pause both in relation to ozone and in terms of small airway research. Work on the air pollution field was about to explode with the demonstration that modern cities were characterized by a statistical association between fine particle pollution (PM2.5) and daily mortality, and that both cardiac and respiratory diseases were responsible for this phenomenon. Dockery & Pope drew attention to this emerging data in their review article in the Annual Review of Public Health, which appeared in 1994 [11]. This association dominated the air pollution field in the 1990’s, and concern about ozone faded into the background. The era of interest in small airways appeared to be over, and Jody Wright edited a very useful summary issue of Seminars in Respiratory Medicine in 1992 which summarized all the work that had been done on small airways. It constitutes another milestone along this road [12]. This volume summarized for the first time the important ozone work being initiated at the Primate Centre at Davis, in Sacramento, which went on to demonstrate the remodeling of the terminal bronchioles that followed exposure of rhesus monkeys to low levels of ozone for the first six months of life.

It seems to be the case with important research questions that they resemble the hydra, in that one head gets chopped off and apparently disposed of, for another to arise and pose a new series of problems. One of the issues that had never been satisfactorily resolved was why repetitive exposures to ozone led to a diminishing response in terms of the decline in FEV1. induced by a single exposure. This meant that with a daily exposure to ozone, by the fourth day, the response had declined almost to zero. Weinmann and her group at Johns Hopkins began to attack this problem in the early 1990’s [13]. They showed that the probable reason for this phenomenon was that the inflammation induced by the first exposure, led to a protective layer of mucus which diminished subsequent responses. They also demonstrated that the effect of ozone on small airways, as indicated by terminal airflow measurements, occurred independently of the C-fibre stimulation [14,15]. By the end of the 1990’s, studies on several hundred normal volunteers exposed to ozone permitted the consequent loss of FEV1 in different concentrations to be precisely defined in relation to symptoms recorded [16], and no further data of this kind were needed. A very important development, pioneered in four laboratories in the United States, was the demonstration that an early effect of ozone was to produce inflammation in the lung, as shown by increased protein and changes in cytokines in bronchial lavage fluid after exposure. These changes were shown to persist for up to 24 hours after the exposure had ceased. All of this work pretty well closed the book on the acute effects of ozone breathing.

At this point there occurred an unusual episode in this rather wayward history that I am following. At the 1990 Air and Waste Management Association meeting, Sherwin presented some histological data on striking small airway changes in the bronchioles of young Los Angeles residents, most of whom had died violently. There were no details of past medical histories. As far as I know, these data never appeared in North American literature, but ten years later, these authors published a comparison between the morphological appearances of the lungs at autopsy of 20 Miami and 18 Los Angeles residents aged 11-30 years. Smoking histories were available for all subjects. The respiratory bronchiolitis was strikingly obvious in the Los Angeles residents and much more severe than the changes in the Miami residents. This analysis was published (in English) in the German journal Virchow’s Archiv [17], and as far as I know has not appeared elsewhere. In 2003, Churg and his colleagues from Vancouver [18] published a comparison between the morphometric differences in the respiratory bronchioles between residents in Vancouver and in Mexico. I annotated this as follows:

Analysis of 20 lungs from women in Mexico City (mean age 66 years)– all never smokers, lifelong residents, no occupational dust exposure, and never used biomass fuel for cooking. These were compared to 20 similar lungs (7 men and 13 women) from Vancouver, BC. Findings were that the Mexico City lungs showed “small airways with fibrotic walls and excess muscle, many containing visible dust. . .” Formal grading analysis (blinded) was conducted. Electron microscopic particle burden on four of the Mexico City cases showed carbonaceous aggregates of ultrafine particles in the airway mucosa. These were considered likely to be combustion products. Aggregates were 0.34-0.54 microns; and the component individual particles were 0.04 – 0.067 microns in size. Muscle hypertrophy was prominent, but there was no basement membrane change. Authors suggest that the demonstrated airway remodeling might lead to chronic airflow obstruction, and in this connection note data from studies of Mexican women which had found an increased risk of chronic bronchitis and chronic airflow obstruction associated with cooking with wood. 3 year mean PM10 in Mexico City is 66 micrograms/m3.

Possible role of other pollutants noted, but authors note similarity of findings to those in occupationally exposed workers to industrial dust who did not have other pollutant exposure. Grading system showed five times higher scores in respiratory bronchioles in Mexico City lungs, and three times higher scores in membranous bronchioles. Excellent and convincing illustrations.

Dr. Churg kindly arranged for me to review these sections, and I was amazed at the difference in the appearances of the bronchioles from the lungs from the two regions – after a little guidance I could say without much hesitation where the sample had come from. The authors attributed the differences to the high particulate pollution in which the Mexico City residents had lived. And of course the ozone levels would have been four or five times higher in Mexico; but the possibility that some or all of the observed changes were due to ozone exposure could not be resolved.

The Southern California Children’s Health study, initiated and conducted by Prof John Peters and his group at the University of Southern California, began to report the results of their detailed comparison of elementary school children in 12 different communities in the Los Angeles Basin. Rather surprisingly, classical cross sectional comparisons between these communities did not show any very striking differences either in symptoms or lung function, although the different communities showed considerable variations in the measured standard air pollutants. In spite of these findings, the research group was able to implement its planned longitudinal component. This programme got over some of the difficulties of cross sectional comparisons, and the first cohort followed for four years showed significant differences in FEV1 and flow rate increases that were related to the ambient concentrations of vehicle emitted pollutants in which the children were living. A second cohort of children had been enrolled immediately after the first, so that it was possible to confirm whether the same phenomenon was demonstrable in them. The results in the first cohort were exactly replicated in the second. This demonstration that the results of a longitudinal study could be exactly replicated in a second cohort, led a high degree of credibility in the conclusions reached. When the results in the two cohorts were combined, they were highly significant, and justifiably made a considerable impression when they were summarized in the New England Journal of Medicine in 2004 [19].

What was surprising was that this phenomenon of slower development of the FEV1 was not related to the ozone level in the different communities (as I had thought it might be) but was clearly influenced only by the package of directly emitted vehicle pollutants and their immediate derivatives. Furthermore, it could be shown that nothing changed if the individual ozone exposures calculated for each child being followed in the study was used, instead of relying only on the ozone measured in the community. Unfortunately we know so little about the factors responsible for normal lung development, that we are far from being able to suggest what specific pollutant or mechanism might be responsible.

The finding that it was the immediately emitted vehicle pollutants that were important, began to be confirmed by a whole series of studies conducted worldwide that indicated that traffic exposure was the important parameter in the associations with acute respiratory outcomes and with increased episodes of bronchitis. The same applied to aggravation of asthma, and possibly to an increased prevalence of asthma, though with this multifactorial disease it proved difficult to pin down the directly relevant pollutant. As far as ozone was concerned, you might have concluded that the main components were now at hand, since although many studies had shown a strong association between this pollutant and hospital admissions for respiratory disease [20], and more recently an associations between ozone and premature mortality [21], there did not seem to be any major differences between communities in relation to the ambient ozone levels in which they lived. The story was not over however.

As you approach the equator, the seasonal variation in ozone which leads to at least a two fold difference in ozone between winter and summer in northern latitudes, begins to disappear. Thus in the city of Brisbane in Queensland, and in Mexico City, ozone levels are fairly constant throughout the year. This makes it possible to study ozone associations without the confounding effect of seasonality, and it is therefore not surprising that the association between asthma hospital admissions and ozone in Brisbane is very strong [20] and that the association between ozone and daily premature mortality is strong in Mexico City [21].

It is at this point that the two themes I have been following, ozone on the one hand and small airway changes on the other, begin to converge (as the earlier calculations of dosimetry indicated they should do).

It became to be realized that the difficulty lay in separating the acute effects of ozone on the FEV1 from the later effects on the small airways. A recent draft of a WHO document on ozone put it this way:

”Ozone produces acute pulmonary inflammation at concentrations near air quality standards. Although the development of tolerance occurs after repeated or chronic exposure to ozone. it is now evident that inflammatory events persist even during the development of functional tolerance to ozone. The persistence of a subclinical inflammatory process may promote permanent damage to or remodeling of pulmonary structure. . . .”

The research team at USC conducting the Southern California Children’s Study started in 1990 had by now developed very strong field teams, and the question was asked as to whether it would be possible to study the prevalence of school absences for acute respiratory illness over a period of several months. Similar studies had been done elsewhere in relation to episodes of pollution, though none of these had included ozone as a featured pollutant. There were many reasons for being dubious as to whether any clear signal could be derived when there were obviously many confounding factors; and the statistical handling of the dat proved to be complex. Nevertheless the group persevered and finally were able to secure data on over two thousand respiratory illness episodes from children in all twelve of the Los Angeles communities, with confirmation in each case that the cause of the absence had indeed been a respiratory illness [24]. The resulting signal, after a formidable result of footwork and refinement of the complex statistical approach to be used, was a strong one. It showed that an increase of 20 ppb in the ambient ozone in the region in which the child was living for 48 hours before the absence, was associated with an increase of 62.9% for illness-related absence rates, 82.9% for respiratory illnesses, 45.1% for upper respiratory illnesses, and 173.9% for lower respiratory illnesses with wet cough. The effects were larger in communities with lower long term PM10 values, but PM10 and NO2 levels were not associated with any increase of risk.

An acute respiratory illness following an ozone exposure might be caused by a number of mechanisms. Ozone interferes with macrophage function for example, and this might intensify the response to a common or garden infective agent, particularly viral in origin. If the levels of ozone induced a degree of small airway inflammation, a subsequent infection might well be more severe. There is therefore plenty of evidence indicating biological plausibility in the case of a “respiratory infection” following a higher ozone exposure. The data from this study formed the basis for an economic estimate of what it represented in terms of dollars [25]. The authors calculated the consequences of the observed reduction in ozone between 1990-1992 and 1997-1999 for rolling three year periods. They found that overall estimates of the number of children’s exposures above 70 ppb on weekdays, from 10 am to 6 pm, indicate a reduction from 83 million exposures per year in 1990-1992 to 17 million per year in 1998-99. The baseline population is the cohort aged 5-18 resident in 1998, and there were 3,283,429 children aged 5-18 in the 1998 population. Between these two intervals, when ozone declined, “the estimated economic value of fewer school absences ranges from $156 million annually to more than $330 million annually, with a best estimate of $245 million. This represents a benefit of nearly $75.00, on average, for every school child in the region”. The overall estimate should be considered conservative, because:

1. Only one day of school absence per episode is assumed:

2. Only values of 8 hour ozone above 70 ppb were taken into account:

3. No estimate was made of medical consultation or medication costs.

This calculation represents a highly credible estimate of the economic burden of current levels of air pollution in one polluted community; and there is no reason to suppose that it would not be operative in any community living in comparable ozone levels.

One of the most difficult challenges is to try and devise ways in which any long term effects of living in a higher oxidant or higher ozone atmosphere, might be studied. Ira Tager, from the Department of Epidemiology at Berkeley University, [26] had the idea some years ago of examining incoming students to that campus with questionnaires and pulmonary function measurements, concentrating particularly on indices of terminal airflow velocity. His first pilot study showed the feasibility of his design, and in 180 students aged about 19 years, all nonsmokers, he found a difference in mean terminal airflow velocity between those who had grown up in high oxidant levels in Southern California, and those who came from Northern California and the Bay area. The lowered airflow velocities were not associated with differences in PM10 or NO2 exposure. He was then funded to repeat the study with a second cohort of entering students, and he sent me the manuscript of this study a couple of days after I had been invited to give this lecture, and it has now been published [24]. Here is my annotation of it:

The study group consisted of 255 incoming students (58% women) to Berkeley who had never smoked, and who had grown up either in the Los Angeles or the Bay areas around San Francisco; mean age about 19. Geocoding of places of residence allowed calculation of average ozone exposure. No associations between any measure of pulmonary function and history of asthma before the age of 12 years, history of pneumonia, bronchitis, allergic conjunctivitis or rhinitis, or ETS exposure. The calculated ozone profile was similar in men and women. Correlations between PM10 and O3 exposure were high in several groups. Results showed: “Consistent inverse associations between increasing lifetime exposures to O3 and FEF75 and FEF25-75 in both men and women”. Associations with NO2 and PM10 were also significant, but reduced substantially when O3 was included.

So my two interests of ozone and small airways have come together. Of course, I cannot prove that what is being induced is a chronic bronchiolitis, but the evidence seems strong enough and consistent enough to me to urge that we must watch the ozone levels to which we are being exposed, extremely carefully. This is made more urgent by the fact that background ozone levels above the Atlantic have been observed to be increasing over the past five years; by the fact that many areas of the United States and Europe now have ozone levels that exceed their current or proposed standards; and that even in the Fraser Valley of British Columbia, to bring it right home, the 8 hour average ozone levels in the summer have been slowly rising over the past ten years. Due to the sharply rising increase in NO2 emissions in Asia, and particularly in China, the background levels of ozone coming to the west coast of America across the Pacific, are predicted to rise slowly but steadily.

I chose to go back over my interest in ozone and small airways because I think we should have a cadre of chest physicians well enough informed about these two converging topics to be able to watch the future scenario with intelligence and perspicacity. My hope is that this record of my interest in this problem will encourage others to become knowledgeable about it.


I am grateful to the Glaxo Company which supported the preparation of this lecture to be given at a Continuing Medical Education course at St. Paul’s Hospital on January 21st 2006. I am also grateful to Dr. Paul Demers of the Occupational Hygiene Institute at the University of British Columbia who allowed me to present it in rehearsal at the combined University of British Columbia and University of Washington Symposium held at Semiahmoo Resort on January 5th 2005.


[1] YOUNG, W.A., SHAW, D.B. & BATES, D.V.
Presence of Ozone in Aircraft Flying at 35,000 Feet.
Aerospace Medicine, 33: 311, 1962.

[2] BATES, D.V., BELL, G.M., BURNHAM, C.D., HAZUCHA, M., MANTHA, J., PENGELLY, L.D. and SILVERMAN, F. Short-term Effects' of Ozone on the Lung.
J. Appl.Physiol. 32: 176, 1972.

[3] BATES, D.V., MILIC-EMILI, J., KANEKO, K., ANTHONISEN, N.R., HENDERSON, J.A.M., DOLLFUSS, R. and DOLOVICH, M. Recent Experimental and Clinical Experience in Studies of Regional Lung Function.
Scand. J. Resp. Dis., Supplement 62, 15, 1966.

Similarity between man and laboratory animals in regional pulmonary deposition of ozone.
Environmental Research 17:84–101, 1978.

A model of the regional uptake of gaseous pollutants in the lung. I. The sensitivity of the uptake of ozone in the human lung to lower respiratory tract secretions and exercise.
Toxicol Appl Pharmacol 79:11–27, 1985.

[6] BATES, D.V. The Respiratory Bronchiole as a Target Organ for the Effects of Dusts and Gases
J. Occup. Med. 15, 177-180, 1973.

[7]. BATES, D.V. The Last Link in the Chain
Amer.Rev. Resp. Dis. 115, 139-141,1977

Proceedings of a Symposium held in Copenhagen, March 29-30, 1979
Excerpta Medica, Amsterdam –Oxford-Princeton. 1979; 259pp.

[9] REED, D., GLASER, S., & KALDOR, J.
Ozone toxicity symptoms among flight attendants
Am J Indust Med 1: 43-54; 1980

Mechanism of action of ozone on the human lung
J. Appl Physiol 67, 1535-1541, 1989

[11] DOCKERY, D.W., & POPE, C.A,.III
Acute Respiratory Effects of Particulate Air Pollution
Annu Rev Public Health, 1994. 15: 107-132

[12] BATES, D.V.
Historical Introduction: The Evolution of Understanding
the Small Airways
In: Seminars in Respiratory Medicine: Edited by J.
Wright. Thieme Medical Publishers Inc, New York, Stuttgart.
Volume 13, Number 2, 63-71, 1992.

Ozone exposure in humans: inflammatory, small and peripheral airway responses
Am J Respir Crit Care Med 152, 1175-1182, 1995

A method for assessing small airways independent of inspiratory capacity
Arch Environ Health 51; 47-51; 1996

Repetitive ozone exposure of young adults
Am J Respir Crit Care Med 164; 1253-1260, 2001

Ozone-induced respiratory symptoms: exposure-response models and association with lung function
Eur Respir J 1999: 14: 845-853

Centriacinar region inflammatory disease in young individuals: a comparative study of Miami and Los Angeles redisents
Virchows Archiv 2000: 437; 422-428

Chronic exposure to high levels of particulate air pollution and small airway remodeling
Environmental Health Perspectives 111: 714-718 (2003)

The Effect of air pollution on lung development from 10 to 18 years of age
New Engl. J. Med 351; 1057-1067, 2004

Associations between outdoor air pollution and hospital admissions in Brisbane, Australia
Arch Environ Health 56; 37-52; 2001

[21] ITO, K., DE LEON, S.F., & LIPPMANN, M.
Associations between ozone and daily mortality
Epidemiology 16: 2005: 446-457

The Effects of ambient Air Pollution on school absenteeism due to respiratory illness
Epidemiology 2001: 12: 43-54

[23] HALL, J.V., BRAJER, V., & LURMANN, F.W.
Economic valuation of ozone-related school absences in the South Coast Air Basin
of California
Contemporary Economic Policy 21; 407-417; 2003

Chronic exposure to ambient Ozone and Lung Function in Young Adults
Epidemiology 16; 2005; 751-759


Copyight Health & Clean Air, 2002-2008