We examined the time course of O3-induced changes in breathing pattern in 97 healthy human subjects (70 men and 27 women). One- to five-minute averages of breathing frequency (fB) and minute ventilation (V̇e) were used to generate plots of cumulative breaths and cumulative exposure volume vs. time and cumulative exposure volume vs. cumulative breaths. Analysis revealed a three-phase response; delay, no response detected; onset, fB began to increase; response, fB stabilized. Regression analysis was used to identify four parameters: time to onset, number of breaths at onset, cumulative inhaled dose of ozone at onset of O3-induced tachypnea, and the percent change in fB. The effect of altering O3 concentration, V̇e, atropine treatment, and indomethacin treatment were examined. We found that the lower the O3 concentration, the greater the number of breaths at onset of tachypnea at a fixed ventilation, whereas number of breaths at onset of tachypnea remains unchanged when V̇e is altered and O3 concentration is fixed. The cumulative inhaled dose of O3 at onset of tachypnea remained constant and showed no relationship with the magnitude of percent change in fB. Atropine did not affect any of the derived parameters, whereas indomethacin did not affect time to onset, number of breaths at onset, or cumulative inhaled dose of O3 at onset of tachypnea but did attenuate percent change in fB. The results are discussed in the context of dose response and intrinsic mechanisms of action.
- breathing pattern
despite the extensive literature examining the effects of the acute inhalation of the photochemical air pollutant ozone (O3) in human subjects, little is known about the time course of O3-induced responses. To a large extent, this deficit is due to the design of the majority of human exposure studies that have been used in establishing the air quality standard for O3. In these studies, pulmonary functions are measured before and immediately following the inhalation of 80–400 parts/billion (ppb) O3 for periods of 0.5–6.6 h in combination with light to severe exercise, while breathing pattern and symptom parameters are evaluated during the initial and final 5 min of exposure. Although studies that examine mechanisms have combined multiple interventions to this design to block or modify one or more O3-induced responses, as a result of these studies, several measurable O3-induced responses have been identified, including symptoms of breathing discomfort, reduced inspiratory capacity, rapid shallow breathing during exercise, mild bronchoconstriction, and airway hyperresponsiveness (33).
O3-induced physiological and symptomatic responses can be viewed as being the result of a complex cascade of events. This cascade is composed of multiple interrelated stages that include but are not limited to 1) the inhalation of O3 and those physical factors that determine the distribution of O3 in the respiratory tract, including airway flow patterns and airway geometry (5, 25, 29); 2) the reaction and uptake of O3 within the respiratory tract, as determined by the composition of the airway lining fluid (ALF), including antioxidants and lipid components (3, 7, 8, 11, 23, 24, 32); 3) the interaction of ozonation products with airway epithelial cells leading to oxidant stress and inflammation, including the release of mediators (15, 16, 34); and 4) the activation of airway sensory nerves by inflammatory mediators initiating pulmonary function decrements, rapid shallow breathing, and symptoms (4, 13, 22, 28, 29, 30). Each stage in this cascade has its own kinetics and gain and potentially contributes to the time course and magnitude of O3-induced physiological and symptom responses. Studies examining the time course of ozone-induced responses would improve the ability to partition different stages of this cascade and determine their relative contribution in determining the magnitude of response.
Anecdotally, participants in O3 inhalation studies report a period of exposure that precedes the onset of symptoms, suggesting a delay in onset of O3-induced response. Further supporting the notion that there is a delay in onset of responses are those studies where O3 exposure has been broken into defined intervals. Schelegle et al. (27) report that O3-sensitive subjects who inhaled 200 ppb O3 in two consecutive 40-min intervals did not have significant alterations in pulmonary function and symptoms after the first 40-min segment but did have significant decrements in function and symptoms after the second segment. Folinsbee et al. (10), McDonnell et al. (19), and Adams (2) have reported that, in 6.6-h exposure protocols broken into six 1-h intervals, the inhalation of 80–120 ppb requires 1–3 h for decrements in pulmonary function to develop. These findings support the existence of a delay in onset of O3-induced physiological responses; however, because of the length of the segments used, it is difficult to precisely define the time required to induce a response at any given inhaled dose rate. The ability to define the time of onset or the cumulative inhaled dose of O3 at onset would greatly improve the ability to predict the minimum no effect dose for O3, as well as potentially providing new insights into underlying mechanisms of O3-induced responses.
In this analysis, we examined the time course of O3-induced increases in breathing frequency (fB). The data used were collected as part of O3-inhalation studies in our laboratory over the last 20 yr (21, 26, 27, 29). All studies were done using continuous exercise that produced constant minute ventilations (V̇e) during the exposure protocols, while the subjects inhaled filtered air (FA) containing O3 by way of a mouthpiece or fasemask that allowed the continuous measurement of fB and tidal volume (Vt). The O3 concentrations used in these studies ranged from 120 to 350 ppb and are in the range of concentrations of other studies that have been used to set national air-quality criteria for O3. This range closely approximates maximum 1-h mean O3 concentrations in the most polluted urban areas in the United States (33). As a reference, the current United States Environmental Protection Agency air-quality standard for O3 is an 8-h average of 80 ppb or 120 ppb for a 1-h average. The United States National Institute for Occupational Safety and Health short-term exposure limit is 300 ppb. We determined the time to onset, number of breaths at onset, and cumulative inhaled dose of O3 at onset of O3-induced increase in fB in 87 normal exercising male and female subjects. Because several of these subjects completed multiple O3-exposure protocols, a total of 248 raw data records were evaluated and used in this analysis. The relationship between the derived time course parameters, individual subject characteristics, and O3-induced decrements in forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1) were examined using stepwise linear regression and correlation analyses. Also available within this data were data subsets that permitted the examination of the effects of altering O3 concentration (350 vs. 200 ppb and 300 vs. 180 ppb), V̇e (70 vs. 50 l/min), atropine treatment, and indomethacin treatment.
The data used in this analysis were derived from the raw data records of six studies that were conducted in the University of California Davis Human Performance Laboratory from 1986 to 2005 and were approved by the University of California Davis, Institutional Review Board. A brief description of each study is given in Table 1. Results from four of these studies have been previously reported in the peer-reviewed literature (21, 26, 27, 29). In total, files were available for 187 healthy human subjects (146 men and 41 women), 18–45 yr of age. Within these files, at least one complete data record for an O3-exposure protocol was available for 87 subjects (67 men and 20 women). As a result of the design of the studies in which they participated, 46 subjects participated in multiple FA and O3-inhalation protocols. As a result, a total of 248 records were evaluated and used in this analysis. Data from each study were considered a data subset. These data subsets were used to examine the effects of O3 concentration (350 vs. 200 ppb and 300 vs. 180 ppb), V̇e (70 vs. 50 l/min), atropine treatment, and indomethacin treatment on the derived end points.
Exposure and measurements.
The same ozone generation and delivery, ventilation monitoring, and pulmonary function equipment was used in all the studies included in this analysis. All equipment was continuously housed, and experiments were performed in the University of California Davis, Human Performance Laboratory, Davis, CA. During each protocol, subjects breathed through a silicone mouthpiece (58 subjects, studies 1–4; see Table 1) or wore a Teflon-coated silicone facemask (29 subjects, studies 5 and 6; see Table 1) attached to a two-way respiratory valve (Hans Rudolph, Kansas City, MO) through which they inhaled FA or O3 in FA while exercising on a cycle ergometer (model 845, Quinton Instrument, Bothell, WA, or model 850, Monarch Exercise, Vansbro, Sweden). To examine whether there was an effect of mouthpiece vs. facemask exposure, we analyzed a subset of the data in which we controlled for total inhaled dose and inhaled dose rate. This analysis showed that there were no significant differences between mouthpiece (n = 38) and facemask (n = 28) for any of the measured end points in this study (comparison not shown).
Air was filtered by a Barneby-Cheney charcoal filter (Columbus, OH) before being inhaled by the subject. Details on the O3-inhalation system are described elsewhere (9). In brief, FA was blended with O3 generated by passing pure oxygen through an ozonizer (Type II, Sander). O3 was continuously sampled from the inspiratory side of the Hans-Rudolph valve and analyzed using a Dasibi O3 monitor (model 1003-AH, Dasibi Environmental, Glendale, CA). The Dasibi monitor was calibrated before and after each study using the ultraviolet absorption photometric method at the California National Primate Research Center, University of California, Davis. In those studies where subjects completed multiple protocols, consecutive O3-exposure protocols were separated by at least 96 h.
Expired air passed through a 5-liter stainless steel mixing and sampling chamber and into a turbotachometer (VMM-2, Interface Associates, Aliso Viejo, CA) that was used to measure expired air flow. To obtain 15-s averages of V̇e, Vt, and fB, the output from the turbotachometer and a temperature thermistor located in the mixing chamber were interfaced with a digital acquisition system.
Atropine and indomethacin treatment.
Fourteen male subjects inhaled atropine sulfate aerosol (0.039 mg/kg, with a range of 0.034–0.046 mg/kg) 20 min before performing preexposure pulmonary function tests. Atropine sulfate aerosols were generated using a modified ultrasonic nebulizer (Mist-O-Gen). The nebulizer was modified so that the expired aerosol was trapped in an attached filter. The amount of atropine delivered was determined by weighing the nebulizer chamber plus filter assembly before and after aerosol inhalation. The dose and time of atropine administration was such that it would result in peak bronchodilation at the time of the preexposure pulmonary function tests and then plateau for the remainder of the time needed to complete a 1-h protocol and postexposure pulmonary function tests. All subjects reported the symptom of dry mouth with atropine inhalation, with 11 subjects also reporting transient dizziness varying in degree from mild to severe. Two subjects also reported mild headache during the atropine inhalation protocols.
Subjects ingested indomethacin (Indocin SR 75, 75 mg of indomethacin, Merck, Sharp, and Dohme, West Point, PA) twice daily with morning and evening meals for 6 days. Subjects inhaled O3 either on day 3 or 6 of this treatment regimen (26).
fB, Vt, and V̇e data were available as 1- or 5-min averages and were transcribed from computer printouts into the database along with pre- and postexposure values for FVC and FEV1 for five of seven studies used. Data stored on computer disks were available for the other two studies. To analyze the time course of fB, Vt, and V̇e responses during a single exposure protocol, the number of breaths and the volume exhaled during each averaging period was calculated. From these data, the cumulative number of breaths and exhaled volume were calculated for each time point. The cumulative number of breaths and exhaled volume were then plotted against time, and cumulative exhaled volume was plotted against cumulative number of breaths. An example of one of these plots is shown in Fig. 1 for a subject who completed a FA and two O3-inhalation protocols (200 and 350 ppb). It should be noted that the slope of the cumulative number of breaths vs. time is equal to fB (Fig. 1A), the slope of the cumulative exhaled volume vs. time is equal to V̇e (Fig. 1B), and the slope of cumulative exhaled volume vs. cumulative number of breaths is equal to Vt (Fig. 1C). After several of these plots were examined, it became apparent that it was possible to detect the time of onset for O3-induced tachypnea and decreasing Vt. It also became apparent that the plot of cumulative breaths vs. time was best approximated by two intersecting straight lines. Time to onset of tachypnea was derived for each O3 inhalation protocol by using regression analysis in combination with visual examination of each plot. In brief, time to onset of tachypnea was determined by using an iterative process involving least squares linear regression (Excel X, Microsoft) of cumulative breaths and time data. In each iteration step, two lines were fitted to the cumulative breaths and time data. In the first iteration, the first region (region 1) of the data that was fitted included the 5-min time point plus the next two time points. The second region (region 2) that was fitted began at the time point 5 min greater than the last time in region 1 and included all the time points greater than this to the end of the protocol. In the next iteration, a time point was added to region 1 and subtracted from region 2. This iterative process was continued until region 2 consisted of the data from the last 5 min of the protocol. With each iteration step, the slope, intercept, and correlation coefficient were calculated for each region. In addition, the difference in slope of regions 1 and 2 and the average correlation coefficient of regions 1 and 2 were calculated. The highest time point value in region 1 where the maximum in the correlation coefficient of region 1, the average correlation coefficient of regions 1 and 2, and the difference in slopes of region 1 and 2 occurred was determined and averaged to obtain the estimated time to onset of tachypnea. This value was then confirmed visually by examining each plot of cumulative breaths vs. time. Those data in which the above criteria were not met by the last iteration and on visual inspection appeared to be a straight line were considered not to have an inflection point and not to contain an onset of tachypnea. These data were then treated similar to the data obtained during FA inhalation (see below).
Once time to onset of tachypnea was determined, an additional six parameters were derived from the data: number of breaths to onset of tachypnea, the cumulative inhaled dose of O3 in micrograms at the onset of tachypnea, the slope before the onset of tachypnea (S1), the slope after the onset of tachypnea (S2), the average Vt before the onset of tachypnea (Vt1), and the average Vt after the onset of tachypnea (Vt2). Five parameters were derived for each FA inhalation protocol: total number of breaths (Nbt), the slope before the last 10 min of the protocol (S1), the slope in the last 10 min of the protocol (S2), the average Vt before the last 10 min of the protocol (Vt1), and the average Vt in the last 10 min of the protocol (Vt2). The percent change in slope (percent change in fB) was calculated for both the FA and O3-inhalation protocols [percent change in fB = 100% × (S2 − S1)/S1]. Similarly, the percent change in FVC, FEV1, FEV1/FVC, and Vt were calculated. In addition, the total inhaled dose rate (in μg/min) was calculated as the product of O3 concentration (in μg/l) times the mean V̇e for each protocol.
Stepwise regression (Statview, SAS) was used to examine the relationship between derived parameters (time to onset of tachypnea and percent change in fB, Vt, FVC, FEV1, and FEV1/FVC), subject characterization (age, height, weight, body surface area, baseline FVC, and baseline FEV1/FVC), and exposure (O3 concentration, O3 inhaled dose rate, total inhaled O3 dose, fB before onset of tachypnea, Vt before onset of tachypnea, and length of exposure) data. Inclusion criteria for independent variables in the regression models was P < 0.05 for each variable’s individual coefficient.
Correlation analysis (Statview, SAS) was used to examine the relationship between the protocol mean values for cumulative inhaled dose of O3 at onset of tachypnea, inhaled dose rate, percent change in fB, and the other derived parameters (percent change in FVC, FEV1, FEV1/FVC, Vt, and time to the onset and number of breaths at onset of tachypnea). Data from the atropine and indomethacin treatment protocols were not included in the stepwise regression or correlation analysis.
Following these analyses, the effect of O3 concentration, V̇e, atropine treatment, and indomethacin treatment was examined in subsets of the database using multivariate ANOVA or ANOVA with repeated measures (Statview, SAS). Specific mean differences were examined using repeated paired t-tests with Bonferoni correction (Statview, SAS). Level of significance was set at P < 0.05 for all comparisons.
Subject characterization data.
All subjects were healthy, nonsmoking, physically active adults without a history of asthma. Subjects that reported having allergic rhinitis were asymptomatic and not on medication at the time of study. Histograms for age (yr), height (cm), body weight (kg), body surface area (m2), baseline FVC (liters) and baseline FEV1/FVC (%) for male and female subjects are shown in Fig. 2. There was no significant difference in age and FEV1/FVC% between male and female subjects. Male subjects were larger on average than female subjects, as indicated by significantly greater height, body weight, body surface area, and baseline FVC in the male subjects compared with the female subjects. Interestingly, there was no significant difference in O3-induced decrements in FVC, FEV1, FEV1/FVC, percent change in Vt, percent change in fB, and cumulative inhaled dose of O3 at onset of tachypnea in comparably exposed male and female subjects. As a result, no further comparison between male and female data were made.
In 20 of the 157 O3 exposure protocols examined an inflection point in fB indicating the onset of tachypnea could not be detected. Most often (19 of 20 cases) this occurred in the protocols utilizing a lower V̇e (50 l/min) and O3 concentration (200 or 180 ppb). Because it was not possible to accurately derive time to, number of breaths at, and cumulative inhaled dose of O3 at onset of tachypnea from these protocols, they were excluded from the statistical analysis. It is interesting to note that in every one of these cases in which an inflection point was not detected the cumulative inhaled dose for the entire protocol was lower than the cumulative inhaled dose of O3 at onset of tachypnea calculated in the same subjects in protocols where an inflection point in fB was detected (Table 2).
Histograms of baseline fB and Vt, as well as V̇e and V̇e/BSA from protocols that were used in the stepwise regression analysis are shown in Fig. 3. Stepwise regression resulted in a significant (P < 0.0001; r = 0.663) association between the time to onset of tachypnea (TOT) and body weight (BWT), O3 concentration ([O3]), fB before onset of tachypnea (fBBOT) and Vt before onset of tachypnea (VtBOT) (Eq. 1; Fig. 4A). Body surface area could be substituted in this relationship without affecting the P value. In addition, there was a significant (P < 0.0001; r = 0.406 and 0.427) association between both the percent change in fB (%fB) and Vt (%Vt) with age, baseline FVC (FVCbaseline), and O3 inhaled dose rate (DR) (Eqs. 2 and 3; Fig. 4, B and C). (1) (2) (3) Interestingly, the percent change in FVC (%FVC) and FEV1 (%FEV1) also resulted in a best fit with equations that included O3 inhaled dose rate (P = 0.0002 and 0.0010, respectively) (Eqs. 4 and 5), whereas percent change in FEV1/FVC (%FEV1/FVC) was best fit with an equation that included O3 concentration (P = 0.0002) (Eq. 6). (4) (5) (6)
Percent change in fB was significantly correlated with percent change in FVC (Fig. 4A), FEV1, and Vt, but not percent change in FEV1/FVC (Table 3). The cumulative inhaled dose of O3 at onset of tachypnea, in contrast, was not correlated with percent change in FVC, FEV1, Vt, and fB, but was correlated with percent change in FEV1/FVC, time to onset of tachypnea, and number of breaths at onset of tachypnea (Table 3). This indicates that there exists some association between the magnitude of the tachypnic response and the pulmonary function responses that are related to O3-induced reductions in inspiratory capacity but not O3-induced bronchoconstriction. In contrast, there was little or no association between cumulative inhaled dose of O3 at onset of tachypnea and the magnitude of pulmonary function responses (percent change in FVC and FEV1), time to onset of tachypnea, number of breaths at onset of tachypnea, inhaled dose rate (Fig. 5A), percent change in Vt, and percent change in fB (Fig. 5B), whereas there was a significant association between cumulative inhaled dose of O3 at onset of tachypnea and percent change in FEV1/FVC (Table 3). In comparison, inhaled dose rate was correlated with percent change in FVC, FEV1, fB (Fig. 5C), Vt, time to onset of tachypnea, and number of breaths at onset of tachypnea but not cumulative inhaled dose of O3 at onset of tachypnea (Fig. 5C) or percent change in FEV1/FVC (Table 3). In addition, percent change in fB was correlated with percent change in FVC (Fig. 5D) and Vt (Table 3).
Two subsets of the data were used to examine the effect of changing O3 concentration on time to onset of tachypnea, number of breaths at onset of tachypnea, cumulative inhaled dose of O3 at onset of tachypnea, and percent change in fB. In the first (study 3 in Table 1), subjects inhaled either 350 or 200 ppb O3 at a V̇e of 50 l/min in two 40-min bouts of exercise separated by a 10-min rest period, during which forced expiratory maneuvers were preformed and a blood sample was collected. Mean values for pulmonary function and fB parameters are given in Table 4. In the second subset, where the effect of O3 concentration was examined (study 4 in Table 1), subjects inhaled either 180 or 300 ppb O3 at a V̇e of 70 l/min for 60 min. In both subsets, inhaling O3 (180, 200, 300, or 350 ppb) resulted in significant decrements in FVC and FEV1 along with significant increases in percent change in fB (P < 0.05). In addition, increasing the inhaled O3 concentration from 180 to 300 ppb or 200 to 350 ppb resulted in significantly greater decrements in FVC and FEV1 and significantly greater increases in fB. The effect of increasing inhaled O3 concentration on FEV1/FVC was more variable (Table 4). Increasing the inhaled O3 concentration resulted in significant decreases in time to onset of tachypnea and number of breaths at onset of tachypnea. In contrast, increasing the inhaled O3 concentration did not result in a significant change in cumulative inhaled dose of O3 at onset of tachypnea (Table 4). Importantly, increasing O3 concentration not only decreased the mean values of time to onset of tachypnea and number of breaths at onset of tachypnea, but also decreased the minimums for these parameters. For example, minimum time to onset of tachypnea decreased from 40.0 to 19.5 min and minimum number of breaths at onset of tachypnea decreased from 1,179 to 511 breaths for the 200 vs. 350 ppb O3 comparison.
A subset of the data (study 4 in Table 1) was used to examine the effect of changing V̇e on time to onset of tachypnea, number of breaths at onset of tachypnea, cumulative inhaled dose of O3 at onset of tachypnea and percent change in fB. In this subset, subjects inhaled 300 ppb O3 at a V̇e of either 50 or 70 l/min for 60 min. Inhaling 300 ppb O3 at a V̇e of either 50 or 70 l/min resulted in significant decrements in FVC, FEV1, and FEV1/FVC along with significant increases in fB (Table 4). Increasing V̇e resulted in significantly greater decrements in FVC and significantly greater increases in fB. In addition, increasing V̇e resulted in a significant decrease in time to onset of tachypnea, but not in number of breaths at onset of tachypnea or cumulative inhaled dose of O3 at onset of tachypnea (Table 4).
Atropine and indomethacin treatment.
A subset of the data was used to examine the effect of the muscarinic receptor antagonist, atropine, or the cyclooxygenase inhibitor, indomethacin, treatment (study 1 in Table 1) on time to onset of tachypnea, number of breaths at onset of tachypnea, cumulative inhaled dose of O3 at onset of tachypnea, and percent change in fB. In this subset, subjects inhaled 350 ppb O3 at a V̇e of 50 l/min for 60 min after receiving no treatment, atropine aerosol, or oral indomethacin. Inhaling 350 ppb O3 with no treatment resulted in significant decrements in FVC, FEV1, and FEV1/FVC, along with significant increases in fB (Table 5). Atropine treatment significantly reduced the O3-induced decrements in FEV1/FVC but did not significantly affect O3-induced decrements in FVC and FEV1 or increases in fB. In contrast, indomethacin treatment significantly reduced the O3-induced decrements in FVC and FEV1 and increases in fB, but did not significantly affect O3-induced decrements in FEV1/FVC. Neither atropine nor indomethacin treatment significantly affected time to onset of tachypnea, number of breaths at onset of tachypnea, or cumulative inhaled dose of O3 at onset of tachypnea. The reported effects of atropine treatment indicate that time to onset of tachypnea, number of breaths at onset of tachypnea, cumulative inhaled dose of O3 at onset of tachypnea, and percent change in Vt and fB are independent of O3-induced bronchoconstriction. The reported effects of indomethacin treatment indicate that time to onset of tachypnea, number of breaths at onset of tachypnea, and cumulative inhaled dose of O3 at onset of tachypnea are independent of the release or production of cyclooxgenase metabolites, whereas percent change in Vt and fB are at least in part dependent on the production of cyclooxgenase metabolites.
This is the first systematic examination of the time course of onset of an O3-induced functional response in human subjects, and it indicates that the inhalation of high ambient concentrations of O3 does not induce an immediate alteration of airway function after one to two breaths but requires many more. The approach used allowed us to determine the time to onset of tachypnea, number of breaths at onset of tachypnea, and cumulative inhaled dose of O3 at onset of tachypnea for each protocol and relate these parameters to the magnitude of response under multiple conditions. These observations improve the understanding of individual responsiveness to O3 and have important implications for exposure models that predict O3-induced responses in the healthy young adult population based on total inhaled dose. Our time course analysis of O3-induced tachypnea revealed a three-phase response. These three phases include a delay phase where no response is detected, an onset phase where fB begins to increase, and a response phase where fB stabilizes at a new elevated level. Selectively changing O3 concentration or V̇e demonstrates that time to onset of tachypnea decreases and percent change in fB increases with increasing inhaled dose rate. This relationship is such that the lower the O3 concentration the greater the number of breaths at onset of tachypnea at a fixed ventilation, whereas number of breaths at onset of tachypnea remains unchanged when V̇e is altered and O3 concentration is fixed. Even more intriguing is the observation that cumulative inhaled dose of O3 at onset of tachypnea remained constant across all exposure conditions and that, if cumulative inhaled dose of O3 at onset of tachypnea was not reached during exposure, functional decrements did not develop. In addition, the observation that once the cumulative inhaled dose of O3 at onset of tachypnea is exceeded the tachypnic response becomes fixed is suggestive that the magnitude of O3-induced tachypnea is more dependent on inhaled dose rate than on the cumulative inhaled dose of O3. Finally, our analysis indicates that the time to onset of tachypnea and the magnitude of the increase in fB are independent and suggest that different intrinsic factors determine each of these parameters.
The majority of studies that have examined O3 dose-response relationships in human subjects have related changes in FVC and FEV1 to the total inhaled dose of O3 or the “effective dose” (1, 14, 31). The obvious question is, do O3-induced decrements in FVC and FEV1 follow the same time course as O3-induced increases in fB? O3-induced decrements in FVC are a result of a decrease in inspiratory capacity, since residual volume is not significantly elevated by O3 inhalation (1, 31, 33). O3-induced decrements in FEV1 are the result of the decreased inspiratory capacity and to a lesser extent mild bronchoconstriction (4, 33) since treatment with atropine in this and another study (4) abolishes O3-induced increases in airway resistance while only producing a mild improvement in FEV1. In turn, reflex inhibition of the ability to inspire has been implicated in both O3-induced decreases in inspiratory capacity (13, 22) and rapid shallow breathing (22, 29). Hazucha et al. (13) found that O3-induced decreases in inspiratory capacity and rapid shallow breathing are not the result of a decrease in dynamic or static lung compliance or respiratory muscle strength (13). Passannante et al. (22) using the opioid-receptor antagonist, sufentanil, were able to reverse O3-induced decrements in FEV1, implicating the involvement of airway C-fibers that are known to be involved in O3-induced rapid shallow breathing in rats (30) and dogs (28). More recently, Schelegle et al. (29), observed that the inhalation of an aerosol of the local anesthetic, tetracaine, significantly reduced and in some cases completely abolished O3-induced symptoms of breathing discomfort, whereas this treatment did not significantly affect O3-induced decreases in inspiratory capacity and rapid shallow breathing. This observation confirms that O3-induced decreases in inspiratory capacity and rapid shallow breathing are independent of symptoms of breathing discomfort and are not the result of a subject’s unwillingness to take complete or deep breaths. Taken together, the findings of these mechanistic studies indicate a common neural origin for O3-induced decreases in inspiratory capacity and rapid shallow breathing that would imply that they would follow a similar time course. This possibility is supported by our observation that, when cumulative inhaled dose of O3 at onset of tachypnea was not achieved, FVC and FEV1 decrements were mild or not present (Table 2).
Another intriguing result of our data analysis was that, once established, O3-induced increases in fB became fixed and were proportional to inhaled dose rate or the product of V̇e times O3 concentration. Again the question to be asked is, do O3-induced decrements in FVC and FEV1 show the same relationship with inhaled dose rate? The result of stepwise regression and correlation analysis on our data set would imply that this is the case. Stepwise regression resulted in inhaled dose rate being included in the best-fit models for O3-induced percent change in fB, Vt, FVC, and FEV1, even when the total inhaled O3 dose (“effective dose”) is included as a possible independent variable in the analysis. Although correlation analysis resulted in significant correlations between percent change in FEV1, percent change in fB, and inhaled dose rate.
It is possible that the apparent increased weight given O3 concentration in predicting FEV1 decrements in previous studies (1, 14, 31) is the result of the fact that, as O3 concentration increases at any fixed V̇e (i.e., increasing dose rate), the number of subjects who would develop FEV1 decrements due to the effect of delay to onset and the magnitude of FEV1 decrements in proportion to dose rate would both increase. As a result, the increasing mean FEV1 decrement with increasing O3 concentration would reflect an increasing number of subjects exhibiting a response, as well as an increased response in those subjects that responded or would have responded at a lower concentration of O3. Consistent with this scenario is McDonnell et al’s. (20) published frequency distribution of percent change in FEV1 as O3 concentration increases.
Three observations are consistent with the notion that the underlying intrinsic factors that determine the magnitude of the delay phase and response phase are independent. First, time to onset of tachypnea, number of breaths at onset of tachypnea, and cumulative inhaled dose of O3 at onset of tachypnea are only poorly correlated, if at all, with percent change in fB, indicating that the length of the delay phase for a given subject does not predict the magnitude of the tachypnic response at any fixed inhaled dose rate. Second, although altering inhaled dose rate by changing O3 concentration or V̇e resulted in related changes in percent change in fB, these alterations did not significantly affect cumulative inhaled dose of O3 at onset of tachypnea. Third, indomethacin treatment, while significantly attenuating percent change in fB, did not change time to onset of tachypnea, number of breaths at onset of tachypnea, or cumulative inhaled dose of O3 at onset of tachypnea. We posit that our observations are consistent with the following cascade of events. As a result of its high reactivity with organic molecules and its low water solubility, O3 on inhalation penetrates into the lower respiratory tract where it reacts rapidly with components of ALF. Initially, O3 reacts with antioxidants such as ascorbic acid, reduced glutathione, and uric acid that are contained in the ALF and act as a defense mechanism against oxidant damage by scavenging free radicals and O3 (3, 7). However, O3 exposure of sufficient duration and concentration can overwhelm these antioxidants, allowing oxidative damage to occur to airway epithelial cells. We propose that this dose of O3 is approximated by the cumulative inhaled dose of O3 at onset of tachypnea.
Once the cumulative inhaled dose of O3 at onset of tachypnea is reached, the reaction between O3 and lipid components of the ALF (fatty acids and surfactant) occurs, producing lipid ozonation products in the form of peroxides and aldehydes that then irritate and/or injure airway epithelial cells (8, 11, 23, 24, 32). Derivatives of ozonized phosphatidylcholine (POPC), the predominant phospholipid found in lung lavage fluid, are capable of activating both cytoplasmic phospholipase-A2 and phospholipase-Cβ (16, 34), which in turn are capable of increasing the availability of arachidonic acid for metabolism, thus producing inflammatory mediators via the cyclooxygenase and lipooxygenase pathways (15, 16). The observation that indomethacin, a cyclooxygenase inhibitor, attenuates O3-induced rapid shallow breathing and decrements in inspiratory capacity but does not affect the cumulative inhaled dose of O3 at onset of tachypnea is consistent with this cascade of events. The intrinsic factors that would determine the magnitude of rapid shallow breathing and decrements in inspiratory capacity would be the amount of LOPs produced, their effect on arachidonic acid release, the activition of cyclooxygenase, and the gain of the neural reflex arcs that results in O3-induced responses (6, 17, 28). Coleridge et al. (17) observed that O3 inhalation in dogs activated both bronchial C fibers and rapidly adapting pulmonary stretch receptors. In turn, lung C fibers have been shown to be the sensory fibers primarily responsible for O3-induced rapid shallow breathing in rats (30) and dogs (28). In humans, neural mechanisms have been implicated in O3-induced symptoms of breathing discomfort (22, 29), decreased inspiratory capacity (13, 22), and rapid shallow breathing (22, 29), as well as bronchoconstriction (4) and airway hyperresponsiveness (12).
The importance of cyclooxygenase product release in the cascade described above is supported by the attenuating effects of indomethacin reported here and elsewhere (26) and the findings of McDonnell et al. (18) who obtained a positive correlation between O3-induced pulmonary function decrements and the level of prostaglandin E2 in bronchoalveolar lavage fluid collected within 1 h after the end of exposure in human subjects who varied greatly in ozone responsiveness. The release of cyclooxygenase products of arachidonic acid from injured airway epithelium can thus be viewed as a link in a cascade of events, which begins with the initial reaction of O3 with the ALF and ends with the observed pulmonary function responses.
Another advantage of the approach used in the present analysis is that by calculating the number of breaths at onset of tachypnea it is possible to better evaluate how breathing pattern and airway geometry affect the onset of O3-induced responses. The factors that determine the absorption of O3 into the ALF, as well as the regional distribution of O3 in the conducting airways during a single breath multiplied by fB, determine the true dose rate of O3 and time course and magnitude of O3-induced responses. During each breath, the fractional absorption of O3 remains relatively constant within a given individual over a twofold change in concentration, V̇e, or time; however, it varies between individual subjects, ranging from 0.80 to 0.91 when these variables are held constant (25). Increasing Vt increases the amount of O3 absorbed into the ALF, and as a result fewer number of breaths are required to obtain a given dose. In addition, increasing Vt will increase the ratio of Vt to anatomical deadspace volume and increase the proportion of the total O3 dose absorbed in the distal conducting airways (5). The increased dose of O3 being absorbed in the distal conducting airways along with our previous observation (29) that O3-induced rapid shallow breathing and reduced inspiratory capacity is initiated by sensory nerves contained within distal conducting airways may account for the trend for cumulative inhaled dose of O3 at onset of tachypnea to decrease as V̇e (and as a result Vt) increases. In addition, the inclusion of body weight or body surface area in the stepwise regression models for the cumulative inhaled dose of O3 at onset of tachypnea is consistent with some dimensional component that is proportional to body size, possibly deadspace volume, contributing to individual differences that determine the dose of O3 needed to induce tachypnea. Further studies need to be conducted in which the deadspace volume, O3 uptake, and time course of O3-induced increases in fB are measured and V̇e is changed while maintaining O3 concentration to definitively address this issue.
The time course analysis of O3-induced rapid shallow breathing in human subjects provides a useful tool to better define the chain of events that lead to O3-induced functional responses, including factors that influence 1) O3 delivery to the tissue (i.e., the inhaled concentration, fB, Vt, and airway geometry); 2) reaction of O3 with antioxidants contained in the ALF; 3) lipid ozonation products inducing local tissue responses, including cyclooxygenase activation; and 4) stimulation of neural afferents and the resulting reflex responses.
This research was funded by California Air Resources Board Contract A4-070-33 and unrestricted gifts from the American Petroleum Institute and Ventaira Pharmaceutical.
The authors thank Emilie Roy for dedication in reviewing all the raw data files, transcribing by hand the data that was available, and performing the initial data analysis. The authors thank Dr. Michelle Fancchi for reading the revised manuscript and providing editorial comments.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society