Vol. 88, Issue 6, 2015-2022, June 2000
Ozone uptake in the intact human respiratory tract:
relationship between inhaled dose and actual dose
Marc L.
Rigas1,3,
Sandra N.
Catlin2,
Abdellaziz
Ben-Jebria3, and
James
S.
Ultman3
1 National Exposure Research Laboratory, U.S.
EPA, Las Vegas 89193-3478; 2 Department of
Mathematical Sciences, University of Nevada Las Vegas, Las Vegas,
Nevada 89154; and 3 Department of Chemical
Engineering, Pennsylvania State University, University Park,
Pennsylvania 16802
 |
ABSTRACT |
Inhaled concentration (C), minute volume
(MV), and exposure duration (T) are factors that may affect the uptake
of ozone (O3) within the respiratory tract. Ten healthy
adult nonsmokers participated in four sessions, inhaling 0.2 or 0.4 ppm
O3 through an oral mask while exercising continuously to
elicit a MV of 20 l/min for 60 min or 40 l/min for 30 min. In each
session, fractional absorption (FA) was determined on a
breath-by-breath basis as the ratio of O3 uptake to the
inhaled O3 dose. The mean ± SD value of FA for all
breaths was 0.86 ± 0.06. Although C, MV, and T all had statistically significant effects on FA (P < 0.0001, P = 0.004, and
P = 0.026, respectively), the magnitudes of these effects were
small compared with intersubject variability. For an average subject, a
0.05 change in FA would require that C change by 1.3 ppm, MV change by
46 l/min, or T change by 1.7 h. It is concluded that inhaled dose is a
reasonable surrogate for the actual dose delivered to a particular
subject during O3 exposures of <2 h, but it is not a
reasonable surrogate when comparisons are made between individuals.
air pollution; breath-by-breath uptake; fractional absorption; exercise
| |
INTRODUCTION |
ACUTE EXPOSURE TO OZONE (O3), a common
outdoor air pollutant, results in impaired lung function such as a
reduced forced expiratory volume and increased airway resistance (10).
Although lung function returns to normal within hours after exposure,
an accompanying tissue inflammation is detectable about 1 day later
(1). The observed increase in lung function decrement associated with
both O3 concentration and the level of physical activity
(13) suggests that this response depends on the dose of O3
delivered to the affected tissues, and the same may be true of airway inflammation.
Because O3 dose to respiratory tract tissue cannot be
directly measured, it is useful to correlate response with a surrogate measure of the actual dose. One logical surrogate is the "inhaled dose," defined as the amount of O3 that is inspired
during a particular duration of exposure. The inhaled dose that occurs
over a series of breaths can be computed as the product of inhaled
O3 concentration (C), inhaled minute volume (MV), and
elapsed time of exposure (T). All previous studies that
attempted to correlate decrements in lung function with the
C · MV · T product have observed
considerable variations between subjects (e.g., Ref. 11). Although
these variations may be due to inherent differences in tissue
sensitivity among individuals, they may also reflect inadequacies in
using inhaled dose as a surrogate for O3 uptake.
The purpose of the present study was to determine O3 uptake
relative to inhaled O3 dose. This ratio, which will be
referred to as fractional absorption (FA), was measured on a
breath-by-breath basis for 10 adult nonsmokers during O3
exposures at alternative C of 0.2 and 0.4 ppm, MV of 20 and 40 l/min,
and total T of 30 and 60 min. To the extent that FA was the same for
all subjects and was independent of C, MV, and T, it could be concluded
that inhaled dose was a reliable surrogate for the actual dose.
| |
METHODS |
Subject characteristics.
Ten healthy adult nonsmokers were recruited for this study.
Prospective subjects underwent a medical screening, including completion of a medical history, physical examination, and forced expiratory spirometry test to determine forced vital capacity (FVC) and
forced expired volume in 1 s (FEV1). Only those individuals who exhibited a FEV1-to-FVC ratio >75% of the predicted
value (6) were included in the study.
Subjects were excluded from the study if they reported a history of
allergies or of any chronic disease, including asthma, allergic
rhinitis, and chronic bronchitis. Subjects were also excluded if they
had smoked or been exposed to urban air pollution regularly within 3 yr
of the screening. Subjects who reported a respiratory illness or the
use of medication (not including contraceptives or vitamin supplements)
within 2 wk of a scheduled experimental session were rescheduled at a
later date. Female subjects were excluded from further study if they
reported that they were pregnant or if positive results were obtained
for a hCG pregnancy test administered immediately before each exposure session.
During the screening session, subjects were given an exercise tolerance
test on a cycle ergometer. Both minute ventilation and heart rate were
monitored as workload was increased in a stepwise fashion to the
subject's maximal level. Each subject was also given a multibreath
nitrogen washout test to determine residual lung volume (RV). Total
lung capacity (TLC) was estimated by the sum of FVC and RV. All
subjects gave informed consent to the screening procedures and the
experimental protocol as approved by the Pennsylvania State University
Institutional Review Board.
The final group of five men (subjects 1-5) and five women
(subjects 6-10) were 18-35 yr old and had the
following mean ± SD characteristics: height = 175 ± 13 cm, weight = 71.6 ± 13.4 kg, FVC = 4.46 ± 1.39 liters, RV = 1.62 ± 0.36 liters, and TLC = 6.11 ± 1.57 liters.
Exposure system.
Ozonated air produced by a commercial O3 generator
(model 03V1-0, OREC, Phoenix, AZ) was diluted with room air
supplied by a 3.5-hp blower (Craftsman Canister vacuum cleaner, Sears,
Chicago, IL) by passing the two streams through a stainless steel
in-line mixer (FMX8413T, Omega Engineering, Stamford, CT). The mixed
stream, flowing at 200 l/min, entered the top of a 30-liter
polycarbonate reservoir and exited through two 1-in.-diameter
respiratory hoses at the bottom of the reservoir. One of these exit
hoses was connected to the inlet port of the subject's breathing
assembly. The other exit hose was vented to a roof exhaust that
continuously removed ozonated air that was in excess of the subject's
respiratory demand while maintaining near atmospheric pressure inside
the reservoir. The O3 concentration in the reservoir was
displayed on a commercial photometric ozone analyzer (1003-AH, Dasibi
Environmental, Glendale, CA) that was used to check the calibration of
a respiratory O3 analyzer and to guide adjustment of the
ozonation system.
Subjects breathed through a low-dead-volume silicone rubber face mask
containing a septum that isolated the mouth and sealed the nose (Series
7900, Hans-Rudolph, Kansas City, MO). As shown in Fig.
1, one end of a pneumotachometer (no. 2, Fleisch, Lausanne, Switzerland) was attached to the oral section of the
mask, and the other end was attached to the common port of a two-way
nonrebreathing valve (model 2700, Hans Rudolph). Subjects inhaled
ozonated air from the 30-liter reservoir through the inlet port of the
nonrebreathing valve and expired into the room through the outlet port
of the valve. The mask-pneumotachometer-valve assembly was supported on
the head with an adjustable harness (Hans-Rudolph) that gave the
subject adequate mobility during the required exercise.
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 1.
Breathing assembly. Inlet of nonrebreathing valve was connected by
flexible tubing to a source of ozonated air and outlet of valve was
open to room air. Sampling port was directly connected to inlet valve
of a respiratory O3 analyzer. Pneumotachometer was
connected by plastic tubing to a differential pressure transducer.
|
|
The inlet sampling line of a custom-built respiratory O3
analyzer (9) was attached to a 1/8-in.-diameter sampling port located
between the exposure mask and the pneumotachometer. When continuously
withdrawing 600 ml/min of air from the respired air stream, the
analyzer had a linear calibration (r2 > 0.99), a
10-90% step response time of <90 ms, a delay time of ~300 ms,
and a minimum detectable limit of 0.006 ppm O3. The analyzer signal was insensitive to changes in temperature, humidity, and carbon dioxide content of respired air.
The two pressure taps on the pneumotachometer were connected by
20-cm-long, 1/4-in.-diameter plastic tubes to a differential pressure
transducer (DP45, Validyne Engineering, Northridge, CA). The voltage
outputs from the respiratory O3 analyzer and the pressure transducer were converted and stored as digital data by a
computer-based data acquisition system (DAS-1601, Keithley Metrabyte,
Cleveland, OH).
Experimental protocol.
At the beginning of each experimental session, the baseline of
the respiratory O3 analyzer was adjusted to zero by
sampling clean air. The sensitivity of the analyzer was checked by
sampling gas from the ozonated air reservoir, the O3
concentration of which was displayed on the Dasibi photometric
analyzer. The dynamic response of the analyzer was also checked by
attaching the sampling line to the common port of a three-way
subminiature solenoid valve (model 4-8-900, General Valve,
Fairfield, NJ) that had one inlet port connected to the ozonated air
reservoir and the other inlet port open to room air. By using a relay
control circuit (PIO-8, Keithley Metrabyte), the data acquisition
system automatically switched the valve inlet position from ozonated
air to room air and then determined 1) the time necessary for
the analyzer signal to begin declining (the delay time) and 2)
the time required for the O3 signal to decline 10% to 90%
between its initial and final levels (the step-response time). The
sensitivity and baseline of the pneumotachometer system were also
checked at the beginning of each session by using a calibrated syringe
to deliver a known volume of air in both the inspiratory and expiratory directions.
Each subject participated in the four exposure sessions listed in Table
1. At the beginning of each session, a
subject donned the face mask and began exercising on a cycle ergometer
(Monark 850, Quinton Instruments, Seattle, WA). The subject
synchronized pedaling frequency to an audible metronome set at 60 beats/min. With the results of the exercise tolerance test for employed
guidance, the torque on the ergometer wheel was adjusted to achieve the desired MV. As soon as the MV of the subject stabilized, the face mask
was connected to the ozonated air reservoir, and the data acquisition
system was initiated. Thereafter, the O3 analyzer and
pneumotachometer outputs were sampled 100 times per second over 30-s
intervals that began 5 min into exposure, 5 min later, and then every
10 min until the end of exposure.
Data computations.
The frequency (f), inhalation time, exhalation time, and period
of the individual breaths obtained within each 30-s sampling interval
were identified from the zero crossing points of the pneumotachometer
signal. The inhaled and exhaled tidal volumes (VT) for each
breath were computed by numerically integrating the pneumotachometer
signal over the inhalation and over the exhalation times, respectively.
The f of a breath was taken to be the reciprocal of the breath period.
Inhaled and exhaled MV were then obtained as the products of inhaled
VT with f and exhaled VT with f, respectively. In effect, each of these two alternative MV represents the ventilation that would occur if an identical breath were repeated for 1 min.
Breath-by-breath calculations of O3 retention required that
the O3 concentration and respiratory flow signals be
properly synchronized. Whereas the flow measurements were assumed to
have a negligible delay time, it was necessary to left-shift the time base of the digitized O3 data according to the delay time
measured for the O3 analyzer. The amount of O3
inspired within an individual breath was determined by numerically
integrating the pneumotachometer signal × time-shifted
O3 signal during the inhalation time, and the amount of
expired O3 was determined by numerically integrating the
same quantity during the exhalation time. The fraction of inhaled
O3 retained in the respiratory system, FA, was then
computed as one minus the exhaled amount of O3 divided by
the inhaled amount of O3.
To analyze the breath-by-breath variation of FA, a database of
"good" breaths was developed. A good breath was defined as having
inspired and expired VT that were within 20% of one
another and as having an inspired VT sufficiently large
that the O3 signal reached a plateau during inhalation. For
each good breath, the database contained the subject identification,
the type of session, the values of f, the inhaled and exhaled
VT, the inhaled and exhaled MV, T and the amounts of
inhaled and exhaled O3, and the value of C, computed as the
average O3 signal near the end of inhalation. The database
contained a total of 2,000 breaths.
Statistical analysis.
Because multiple measurements were obtained from individuals at
different times, we adopted a multivariate approach with structured covariance for individuals to statistically analyze the FA data in
terms of the three factors C, MV, and T. The analysis was based on a
linear mixed-effects model (7) given by
where
j denotes the subject and k the observation on the
subject. This is a type of two-stage model in which the fixed effects (
i, i = 0-3) can be considered as
the first stage and the random effects (
ij,
i = 0-3) that vary by subject as the second stage. It was
assumed that the components of j 
(0j,
1j,
2j,
3j)T were independent
and identically distributed N(0,
) where is a 3 × 3 covariance matrix for j. It was further
assumed that the 
jk were independent and
identically distributed N(0,
2) and were independent of
j. Because each measurement period was
only 30 s, it was reasonable that each observation would be independent
because any effect of time would not be noticeable over this interval.
Moreover, statistical tests on observations during these intervals
showed that autocorrelation was not significant.
The function lme in the statistical package Splus (MathSoft,
Seattle, WA) was used to fit the linear mixed-effects model. The
parameters in the variance structure were estimated by first maximizing
the marginal likelihood of the residuals obtained by the least-squares
fit. The fixed effects were then estimated via maximum likelihood
assuming the variance structure to be known, such as in a generalized
least-squares procedure. Computational procedures are detailed in
Lindstrom and Bates (8). Approximate standard errors for
the fixed effects were derived via asymptotic theory (12).
In addition to running the model for MV and C as continuous variables,
we used the target values of these variables (Table 1) for each
exposure condition and analyzed them as fixed factors. These fixed
factors will be referred to as mv and c.
| |
RESULTS |
Continuous data recordings.
The raw data taken during a typical series of breaths are shown
in Fig. 2, in which the vertical dotted
lines indicate the beginning of inspiration. Figure 2A is the
calibrated output from the O3 analyzer that was shifted to
the left by 28 sampling intervals to account for the 280-ms time delay
observed at the beginning of the session. Figure 2B is the
calibrated signal from the pneumotachometer. In Fig. 2C, the
pneumotachometer signal has been numerically integrated to depict
respired volume.
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 2.
Breath-by-breath data for subject 4 during exposure session
3. Ozone analyzer output after low-pass filtering at 25 Hz
(A), pneumotachometer output (B), and integrated
pneumotachometer output (C) are shown. Dotted lines indicate
beginning of an inhalation.
|
|
The O3 concentration observed near the end of all breaths
dropped to zero or even slightly below zero (Fig. 2A). The
negative values of the analyzer output that appeared in some breaths
were probably due to a slow thermal drift in the electronics of the O3 analyzer. To eliminate the small effect that this
artifact could have on the computations of O3 retention,
all the O3 data in a particular breath were shifted upward
by a constant that forced the end-expired O3 concentration
to be zero.
Breathing patterns.
Inspired and expired breathing parameters were not the same.
The average value of the inhaled MV for all 2,000 breaths in the
database was 28.8 ± 0.2 l/min (mean ± SD), whereas the average expired MV was 30.5 ± 0.3 l/min. The average inhaled VT
was 1.10 ± 0.01 liters, whereas the average exhaled VT
was 1.17 ± 0.01 liters. The fact that VT and MV were
~7% larger during exhalation than during inhalation was probably due
to the heating and humidification of inhaled air within the respiratory
system. In carrying out all the statistical analyses that follow, the
inspired MV rather than the expired MV was used as a factor. The
expired MV was disregarded for two reasons: first, because the
pneumotachometer was calibrated with room air, inspiratory flow
measurements were more reliable than expiratory flow measurements; and
second, because the exhaled O3 concentration always reached
zero well before the expiration ended, the exhaled MV could not have
had an important influence on FA.
Table 1 indicates that the MV of 20 and 40 l/min that were targeted by
employing light and moderate exercise challenges, respectively,
occurred primarily by differences in VT rather than by
differences in f. Moreover, an analysis of all 2,000 breaths in the
database indicated that MV was clearly associated with VT
(Pearson's correlation coefficient: r2 = 0.44),
whereas MV was essentially uncorrelated with f (r2 = 0.01). This result is consistent with previous reports for subjects
exercising within their aerobic range (5) and implies that MV was
directly proportional to VT and f was relatively constant.
O3 retention.
The values of FA ranged from 0.56 to 0.98 with a mean ± SD of
0.86 ± 0.06 for all 2,000 breaths. Figures
3, 4, and
5 examine how c, mv, and T influenced FA on
a subject-to-subject basis. Although it is not clear in Figs. 3 and 5
whether there was a consistent change in FA with c or with T, Fig. 4
illustrates that FA increased with mv for most subjects. By comparing
pooled FA data at the different values of c, mv, and total T as well as for the different subjects, Fig. 6
demonstrates that intersubject differences had the most influence on
FA, causing a variation of ~10%.
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 3.
Effect of target inhaled concentration (c) on the fractional absorption
(FA) from individual subjects. Subject-by-subject box plots indicate
distribution of FA among breaths recorded during sessions 1 and
3 (c = 0.2 ppm) and during sessions 2 and 4 (c = 0.4 ppm). Box represents range of middle 50% of data. Unfilled
rectangle inside box is median. Whiskers represent 1.5 times
interquartile range, and outliers are indicated by solid horizontal
lines. FA distributions for subject 1 do not include
sessions 2 and 3 and for subject 8 do not
include sessions 3 and 4.
|
|
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 4.
Effect of target inhaled minute volume (mv) on FA from individual
subjects. Subject-by-subject box plots indicate distribution of FA
during sessions 3 and 4 (mv = 20 l/min) and during
sessions 1 and 2 (mv = 40 l/min). Box represents range
of middle 50% of data. Unfilled rectangle inside box is median.
Whiskers represent 1.5 times interquartile range, and outliers are
indicated by solid horizontal lines. FA distributions for subject
1 do not include sessions 2 and 3 and for
subject 8 do not include sessions 3 and 4.
|
|
View larger version (32K):
[in this window]
[in a new window]
|
Fig. 5.
Effect of exposure time (T) on FA from individual subjects.
Subject-by-subject box plots indicate distribution of FA at T when data
were recorded. Box represents range of middle 50% of data. Unfilled
rectangle inside box is median. Whiskers represent 1.5 times
interquartile range, and outliers are indicated by solid horizontal
lines. FA distributions for subject 1 do not include
sessions 2 and 3 and for subject 8 do not
include sessions 3 and 4.
|
|
View larger version (10K):
[in this window]
[in a new window]
|
Fig. 6.
Relative effects of all factors on FA. This plot depicts mean value of
FA at each level of factors subject (Sbjct), total exposure duration
(Time), mv, and c.
|
|
In the statistical analyses, we elected to include all the measurements
from the four sessions, even though data was collected for 30 min more
in light exercise sessions 3 and 4 than in moderate exercise sessions 1 and 2. In essence, it was assumed
that the exercise effect is linear throughout the exposure period. In
fact, the model and conclusions were virtually unchanged when the 30 min of extra data were omitted or when data from the 10- to 30-min intervals were compared with data from the 40- to 60-min intervals in
sessions 3 and 4. Because subject 1 did not
participate in sessions 2 and 3 and subject 8 did not participate in sessions 3 and 4, it was not
possible to estimate random effects for these individuals, and their
data were not used in the statistical analysis. A total of 1,833 breaths from eight subjects remained in the database.
The fit of the mixed-effects model to the data was significantly better
(via Akaike information criterion, Bayesian information criterion, or
likelihood ratio criteria) when
3j was omitted than when this random
effect of inhaled concentration was included in the model. After
3j was excluded, the results of the
model applied to the continuous factors indicated that MV
(P = 0.0011), T (P = 0.004), and C (P < 0.0001) all had fixed effects that were significantly different
from zero. The fixed effects were estimated to be 0 = +0.865, 1 = +0.00110 min/l, 2 = 
0.000499 min
1, and
3 = 0.0385 ppm1.
In general, the random effects, 1j
and 2j, were an order of magnitude
smaller than the corresponding fixed effects, 1 and
2. The net effect of MV on FA, as judged by
1 + 1j, was positive
in all except one subject. For subject 9, a negative
1 + 1,9 value of 0.00114 min/l
reduced the significance of the positive overall influence of MV. The
net effect of T on FA, as judged by 2 + 2j, was negative in all except one
subject. For subject 3, the positive 2 + 2,3 value of +0.000269 min1 reduced the significance of the
negative overall influence of T.
The same mixed-effects model was also applied to the fixed-factor
levels, c and mv. Consistent with the analysis employing continuous
factors, mv (P = 0.005) and c (P < 0.0001) exhibited fixed effects that were significantly different from zero. Unlike the
previous analysis, however, T (P = 0.15) was not significant. The estimates of fixed effects were 0 = 0.891, 1 = +0.0134 min/l, 2 = 0.000261
min1, and 3 = 0.0613 ppm1, and random effects
were of the same order of magnitude as the corresponding fixed effects.
Subject 9 was again the only individual to exhibit a negative
value of 0.0136 min/l for 1 + 1,j, but subjects 2 and 3 had positive estimates for 2 + 2j of +0.000165 and +0.000474
min1, respectively.
| |
DISCUSSION |
In this study, the fraction of inhaled O3 retained in the
respiratory system during a single breath was determined by integrating O3 concentration and respiratory flow data monitored in 10 healthy adult nonsmokers who were orally exposed to either 0.2 or 0.4 ppm O3 for 30 or 60 min while exercising at light or
moderate workloads designed to elicit MVs of 20 or 40 l/min. Because
expired O3 concentration reached zero before the end of
each breath, there was no O3 buildup in the gas spaces of
the respiratory tract, and (inhaled dose) · FA was
equivalent to the amount of O3 absorbed into the epithelial
lining fluid during the course of a breath. Because O3
rapidly reacts in this protective fluid layer, the amount of
O3 that reaches the underlying respiratory tract tissue is
less than (inhaled dose) · FA. Even so, the
measurement and analysis of FA bring us one step closer to knowing
tissue dose than the measurement of inhaled dose alone.
The statistical analyses carried out in this study indicated that MV,
T, and C all had significant effects on FA. The estimates of the fixed
effects, however, suggested that the influence of MV, T, and C may not
always be important from a practical point of view. For example,
3 was estimated as 0.0385
ppm1, so C would have to change by
0.05/0.0385 = 1.3 ppm to induce a 5% change in FA. Because changes in
C of this magnitude are highly unlikely in occupational as well as
recreational settings, the influence of C on FA is clearly unimportant.
By similar reasoning, MV would have to undergo a change of 46 l/min for
FA to change by 5%. This might happen if an individual drastically
changed his or her level of physical activity between rest and heavy
exercise. On the other hand, a change in T of only 100 min would be
required to change FA by 5%. This condition is often met in both
natural and controlled laboratory exposures.
The FA predicted from the statistical model using mv, T, and c as
factors were highly correlated with the FA predicted from the model
using MV, T, and C as factors (r2 = 0.957). This
suggests that intersubject variability contributed more to the
variability in FA than the inaccuracy of approximating MV and C
by their target levels. Figure 6 further shows that differences in FA
among subjects were much greater than FA differences caused by changes
in mv, c, and T. This may explain why, in previous O3
exposure studies that used inhaled dose or inhaled dose rate as
surrogates for actual uptake, the response data from individual subjects were so scattered in the dose-response plots (e.g., Ref. 11).
The positive relationship between FA and MV revealed by the statistical
analyses was probably a consequence of the fact that increases in MV
were accomplished more as a result of an increase in VT
than an increase in f. An increase in VT implies that
O3 penetrates deeper into the respiratory tract, in which
absorption is a more efficient process (4). The finding that FA
decreased as T increased may have been due to a progressive depletion
of reactive substrates in the liquid lining layer (2). This would result in a buildup of O3 "backpressure" that reduced
the concentration driving force for absorption.
Only a few measurements of overall respiratory O3 uptake
have previously been reported. Employing a pneumotachometer and a rapidly-responding chemiluminescent O3 analyzer similar to
the instrument used in the present study, Gerrity et al. (3) determined FA using essentially the same breath-by-breath O3 retention
calculations as those employed in the present research. For 10 healthy
adults engaged in quiet oral breathing at a C of 0.4 ppm
O3, T of ~60 min, and an average MV of 9.6 l/min, the
average ± SE value of FA was reported to be 0.907 ± 0.010. This
compares favorably to the average ± SD value of 0.86 ± 0.06 for the
2,000 breaths analyzed in the present study.
Wiester et al. (14) measured average uptake into the lungs of 10 healthy men who breathed quietly from a mask affixed to a
large-diameter pipe through which 0.3 ppm ozonated air was supplied at
a flow of 40 l/min. A mixing chamber located downstream of the subject
was used to dampen fluctuations in O3 concentration. With
this apparatus, O3 uptake could be determined by
multiplying the steady upstream-to-downstream decrease in
O3 concentration with the ozonated air flow entering the
pipe. This experimental design avoided the need for a fast-responding
instrument so that a slowly-responding commercial O3
analyzer could be utilized. For quiet oral breathing at a C of 0.3 ppm
O3, a total T of ~30 min, and an average MV of 10.4 l/min, the average value of FA was reported to be 0.765. This is less
than the average value of 0.850 predicted for these breathing
conditions by using the fixed-effect estimates of the mixed-effects model.
Another discrepancy is the conclusion by Wiester and associates (14)
that FA was a decreasing function of MV (their Fig. 4) rather than an
increasing function of MV, as found in the present study. Their
conclusion was based on FA values that ranged from 0.96 to 0.51 as MV
increased from 8 to 12 liters. This is in contrast to the present study
in which the FA measured in 2,000 individual breaths ranged from 0.56 to 0.98 as MV increased from 10 to 70 l/min. It is possible that there
was a difference in breathing patterns of the subjects in the two
studies; the subjects in the present study tended to increase MV by
increasing VT, whereas the subjects in the previous study
may have increased f as well as VT. In all probability,
however, the conclusion of Wiester and associates was an artifact of a
large intersubject variation occurring over a comparatively narrow
range of MV. Moreover, the precision of the uptake measurements of
Wiester and associates is somewhat suspect because they reported up to
25% variability in uptake within one subject between two different
laboratory visits. The session-by-session variability in the present
study was 6% at most.
In conclusion, when a particular subject is exposed to a fixed
concentration of O3 at a fixed exercise level, inhaled dose is a reasonable surrogate for the actual uptake of O3 by
the respiratory system, as long as the T is <2 h. On the other hand,
the actual dose may vary considerably among individuals who are exposed
to similar inhaled doses.
| |
ACKNOWLEDGEMENTS |
The U.S. Environmental Protection Agency through its Office of
Research and Development collaborated in the research described here.
This research has been subjected to Agency review and approved for
publication. Mention of trade names or commercial products does not
constitute an endorsement or recommendation for use.
| |
FOOTNOTES |
This work was partially funded by National Institute of Environmental
Health Sciences research grant RO1 ES-06075. Clinical support was
provided by the General Clinical Research Center of the Pennsylvania
State University through funding by National Center for Research
Resources grant MO1 RR-10732.
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. §1734 solely to indicate this fact.
Address for reprint requests and other correspondence: J. S. Ultman,
Dept. of Chemical Engineering, Pennsylvania State Univ., 106 Fenske
Laboratory, Univ. Park, PA 16802 (E-mail: jsu{at}psu.edu).
Received 21 October 1998; accepted in final form 8 February 2000.
| |
REFERENCES |
1.
Aris, RM,
Christian D,
Hearne PQ,
Kerr K,
Finkbeiner WE,
and
Balmes JR.
Ozone-induced airway inflammation in human subjects as determined by airway lavage and biopsy.
Am Rev Respir Dis
148:
1363-1372,
1993[ISI][Medline].
2.
Asplund, PT,
Ben-Jebria A,
Rigas ML,
and
Ultman JS.
Longitudinal distribution of ozone absorption in the lung: effect of continuous inhalation exposure.
Arch Environ Health
51:
431-438,
1996[ISI][Medline].
3.
Gerrity, TR,
Biscardi F,
Strong A,
Garlinton AR,
Brown JS,
and
Bromberg PA.
Bronchoscopic determination of ozone uptake in humans.
J Appl Physiol
79:
852-860,
1995[Abstract/Free Full Text].
4.
Hu, SC,
Ben-Jebria A,
and
Ultman JS.
Longitudinal distribution of ozone absorption in the lung: effects of respiratory flow.
J Appl Physiol
77:
574-583,
1994[Abstract/Free Full Text].
5.
Johnson, AT.
Biomechanics and Exercise Physiology. New York: Wiley, 1991, p. 314.
6.
Knudsen, RJ,
Slatin RC,
Lebowitz MD,
and
Burrows B.
The maximal expiratory flow-volume curve.
Am Rev Respir Dis
113:
587-600,
1976[ISI][Medline].
7.
Laird, NM,
and
Ware JH.
Random-effects models for longitudinal data.
Biometrics
38:
963-974,
1982[ISI][Medline].
8.
Lindstrom, MJ,
and
Bates DM.
Newton-Raphson and EM algorithms for linear mixed-effects models for repeated-measures data.
J Am Stat Assoc
83:
1014-1022,
1988[ISI].
9.
MacDougal, CS,
Rigas ML,
Ben-Jebria A,
and
Ultman JS.
A respiratory ozone analyzer optimized for high resolution and swift dynamic response during exercise conditions.
Arch Environ Health
53:
161-174,
1998[ISI][Medline].
10.
McDonnell, WF,
Horstman DH,
Hazucha MJ,
Seal E,
Haak ED,
Salaam SA,
and
House DE.
Pulmonary effects of ozone exposure during exercise: dose-response characteristics.
J Appl Physiol
54:
1345-1352,
1983[Abstract/Free Full Text].
11.
McDonnell, WF,
Stewart PW,
Andreoni S,
Seal E,
Kehrl HR,
Horstman DH,
Folinsbee LJ,
and
Smith MV.
Prediction of ozone-induced FEV1 changes.
Am J Respir Crit Care Med
156:
715-722,
1997[Abstract/Free Full Text].
12.
Pinheiro, JC,
and
Bates DM.
Unconstrained parametrizations for variance-covariance matrices.
Stat Comput
6:
289-296,
1996.
13.
Tilton, BE.
Health effects of tropospheric ozone.
Environ Sci Technol
23:
257-263,
1989.
14.
Wiester, MJ,
Stevens MA,
Menache MG,
McKee JL,
and
Gerrity TR.
Ozone uptake in healthy adult males during quiet breathing.
Fundam Appl Toxicol
29:
102-109,
1996[Medline].
J APPL PHYSIOL 88(6):2015-2022
8570-7587/00 $5.00
Copyright © 2000 the American Physiological Society
TOP
ABSTRACT
INTRODUCTION
