United States
Environmental Protection
Agency
Health Effects Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S1-81-022 May 1981
Project Summary
Effects of Etiologically Defined
Respiratory Infections on Lung
Function and Its Growth in an
Area of Low Air Pollution -
Spirometry in Young Children
When Illness-Free
Albert M. Collier, Gerald L. Strope, Ronald W. Helms, Lisa Morrisey Lavange,
Wallace A. Clyde, Jr., Russel Pimmel, Floyd W. Denny, Frederick W.
Henderson, and Gerald W. Fernald
This longitudinal study was per-
formed in a group of 3-12 year old
children to document normal lung
growth patterns as measured by spi-
rometry. By clinical and laboratory
parameters, these children were free
of illness at the time of study and had
been for the preceding 21 days. Spiro-
metry was performed prospectively
over a period of six years in 69 children
(27 black females, 23 black males, 10
white females, and 9 white males)
from a day care center. Eight-hundred
fifteen spirometric tests were made on
these children. Linear regres-
sion analyis by a method of weighted
least squares was found to adequately
describe the data over a height range
of 100-150 cm and was performed on
six spirometric parameters: forced
vital capacity (FVC), forced expiratory
volume in one second (FEVi), peak
expiratory flow (PEF), forced expira-
tory flow during the middle half of the
FVC (FEF25-75%), and maximum expira-
ory flows after 50% and 75% of the
FVC have been exhaled (Vmax5o% and
Vmax75%, respectively). Ninety-five
percent confidence limits for the
regression lines and 95% prediction
intervals for individual observations
were also computed. There were sig-
nificant differences between the re-
gression lines (considering slope and
intercept) for all six parameters when
black females were compared to black
males, white females to white males,
black females to white females and
black males to white males, except for
Vmax75% for the comparison of black
females to black males and white
females to white males. These slopes
and intercepts were similar to those
reported by others for children of
similar age and height. The 95% confi-
dence limits and 95% prediction inter-
vals were proportional to height and
were similar to estimates of parameter
variability reported by others. This
study demonstrates that spirometry
can be performed reliably at an early
age in a day care center population,
that there are significant racial and
sexual differences in spirometric vol-
ume and flow parameters and that the
variability of these measurements is
-------
proportional to height rather than
being a constant.
This final report was submitted in
fulfillment of Grant #R-804577 by
the University of North Carolina under
the sponsorship of the U.S. Environ-
mental Protection Agency. This report
covers the period from August 1,
1978 to April 30, 1980 and was
completed as of April 30, 1980.
This Project Summary was devel-
oped by EPA's Health Effects Research
Laboratory. Research Triangle Park,
NC. to announce key findings of the
research project that is fully docu-
mented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
The growth of the lung in children isa
dynamic process characterized by an
increase in alveolar, number followed by
an increase in size of airways and air
spaces in the lung. Other researchers
have shown that the total number of
airways is the same in children and
adults. The total complement of bronchi
al generations is present by the end of
the sixteenth week of gestation and the
spaces which will contain air begin to
change to an alveolar pattern during the
latter part of gestation. During the first
weeks of life, typical alveoli appear and
lung growth for the first four years of life
consists predominantly of an increase
in alveolar number. At this point, increase
in lung size changes to a pattern of
growth consisting of an increase in size
of airways and alveoli. The physiologic
importance of this growth pattern has
been demonstrated by other researchers
who found that the conductance of
peripheral airways (beyond the fifteenth
bronchial generation) increases dra-
matically at about five years of age,
while no change occurs in the central
airways.
Most reference data for pulmonary
function tests in children have been
generated on newborns and on children
over five or six years of age. Pulmonary
function tests from infancy to five years
of age have not been well documented
and it is not known what effect the
relative lower conductance of the pe-
ripheral airways might have on measure-
ments of lung function. It is also not
known what effects acute respiratory
illnesses might have on measurements
of pulmonary function and if these
effects are more easily detected because
conductance of the peripheral airways
is lower. Previous research has shown
that the peripheral fraction of total
respiratory resistance measured by the
forced random noise technique in a
group of children, whose average age
was 46 months, to be significantly
greater than in a group of children
whose average age was 61 months.
(Conductance of the peripheral airways
was less in younger children than in the
older children.) Other researchers have
demonstrated a greater number of
significantly decreased spirometric
parameters during acute upper respira-
tory illness in a group of children less
than 84 months of age, as compared to a
group of children greater than 84 months
of age. These data suggest that there
are differences in lung function during
early life attributable to developmental
phenomena in the airways and that
these differences may be important in
the manifestation of disease.
Additionally, data of pulmonary func-
tion in children have been collected in a
cross-sectional fashion. No data exist
which have been systematically gener-
ated in a longitudinal fashion. These
data are important to determine if lung
function development in children pro-
gresses in a predictable fashion much
the same as height and weight develop-
ment, or if the growth of the lung might
progress in another as yet undetermined
pattern.
This report outlines theanalysisof the
spirometry data collected over a six-year
period at the Frank Porter Graham Child
Development Center. Only information
gathered on the children when they
were well has been included in this
analysis. The data have been analyzed
in a cross-sectional fashion so that later
analyses can beperformedtodetermine
what the relationships of each individual
child's growth pattern is to the overall
growth pattern. Because of the known
effects of race and sex on lung function,
the children have been separated into
four groups for these analyses: black
females and males and white females
and males.
Conclusions
Over the six year study period, 69
children were followed for an average of
3.5 years each and had an average of
3.14 measurements made per year. The
children, in general, were 90 to 155cm.
tall and spanned an age range of 2'/2 to
12 years. Only data from the children
whose heights were between 100 and
150 cm., however, were included in the
regression analyses.
The regression coefficients and stan-
dard errors of the six spirometric param-
eters computed by weighted least squares
analysis for the four race/sex groups
are shown in Table 1. The coefficients
for the 95% confidence and 95% predic-
tion limits are also listed. Plots of each
discrete data point along with the
regression lines and 95% confidence
and prediction intervals for each param-
eter in each of the four race/sex groups
reveal an excellent fit of the data to the
regression lines. Comparisons of plots
of the regressions for the four race/sex
groups in general reveal that, for any
given height, white children have larger
lung function parameters than black
children. The p-values for the tests of no
difference comparing the regression
lines (including slope and intercept)
between the important race/sex group-
ings are shown in Table 2. All of these
comparisons are significantly different
except for v"max75% for black females
compared to black males and for white
females compared to white males. The
variability parameters (95% prediction
limits) as computed in these analyses
are similar to variability as computed by
more classical methods (e.g.. Mean ±
2SD). This suggests that making mea-
surements in a longitudinal fashion
does not reduce the expected variability,
as demonstrated below.
1. The feasibility of obtaining reproduc-
ible spirometric tests of lung function
in small children as young as three
years of age.
2. Significant racial and sexual differ-
ences in most spirometric parameters
of lung function in young children.
3. That regression lines of spirometric
parameters of lung function can be
adequately described in young chil-
dren by a linear regression over the
height range of 100-150 cm.
4. That parameters of variability are
proportional to height rather than
being a constant over the region of
interest.
Recommendations
To ascertain the effects of acquired
inslut on the lung such as from air
pollution on infectious agents, additional
reference data from populations studied
in an area with minimal air pollution,
and from populations whose illness
patterns have been closely documented,
are needed. Until the effects of lower
-------
Table 1. Weighted Regression Coefficients* for Spirometric Parameters
Black Females
Black Males
Parameter*
FVC. '
FEVi, '
PEF. '/sec
FEFzs-75%. '/sec
VmaxsoVo, '/sec
Vmax7s%, '/sec
b"
-1.966±0.084
-1.533±0.076
-3. 1 12±0.290
-0.493±0.211
-0.820±0.245
0.243±0.177
m{
0.0267±0.0008
0.0222±0.0007
0.0531 ±0.0026
0.0195±0.0019
0.0245±0.0022
0.007 3±0.00 16
K*
(xlO3)
0.415
0.378
1.477
1.041
1.213
0.865
Lc
(xlO2)
0.371
0.334
1.281
0.931
1.084
0.781
b
-2.601 ±0.1 00
-2. 191 ±0.090
-4.913±0.343
-1.650±0.246
-1.888±0.291
-0.297±0.209
m
0.0324±0.0009
0.0281 ±0.0008
0.0685±0.0031
0.0292±0.0022
0.0330±0.0026
0.0119±0.0019
K
(xlO3)
0.485
0.437
1.692
1.231
1.478
1.032
L
(x102)
0.372
0.335
1.284
0.933
1.087
0.783
White Females
White Males
Parameter
K
(xlO3)
L
fx102)
m
K
(xlO2)
L
(x103)
FVC, '
FEI/,, '
PEF. '/sec
FEFZ5-7s%, '/sec
VmaxsoVo, '/sec
1/073*75%, '/sec
-2.969±0.165 0.0365±0.0013 0.708 0.374 -2.442±0.152
-2.469±0.148 0.0313±0.0012 0.639 0.336 -1.676±0.136
-3.846±0.569 0.0611 ±0.0046 2.393 1.288 -1.378±0.523
-1.644±0.414 0.0316±0.0033 1.739 0.936 -0.077±0.380
-2.111 ±0.482 0.0380±0.0039 2.026 1.090 -0.287 ±0.443
-0.256±0.347 0.0130±0.0028 1.485 0.786 0.466±0.319
0.0331±0.0013 0.663 0.374
0.0256±0.0011 0.581 0.336
0.0402±0.0043 2.227 1.288
0.0174±0.0032 1.618 0.937
0.0209±O.O037 1.885 1.091
0.0067±O.O026 1.390 0.786
"Regression equations have the form y=mx + b, where y represents the parameters; m. the slope with units equal to those of the
parameter divided by cm; b, the intercept with units identical to the parameter; and x, the height in cm.
"Ninety-five percent confidence limits (for the regression line) for any height between 100 and 150 cm can be computed using the
form (mx + b) ± Kx. where K represents the 95% confidence coefficient.
cNinety-five percent prediction limits (for an individual observation) for any height between 100 and 150 cm can be computed using
the form (mx + b) ± Lx, where L represents the 95% prediction coefficient.
"Parameters are FVC (forced vital capacity). FEVi (forced expiratory volume in one second), PEF (peak expiratory flow). FEFzs-7s°/o
(forced expiratory flow during the middle half of the FVC), Vmax50%, (maximum expiratory flow after 50% of the FVC has been
exhaled), and Vmax7s% (maximum expiratory flow after 75% of the FVC has been exhaled).
"b - intercept ± S.E.
'm = slope ± S.E.
respiratory illnesses during infancy on
lung growth and development are better
understood, populations whose entire
life history of respiratory illness is
known should be studied. Since lower
respiratory illnesses during infancy may
be related to altered lung function later
in life, these illnesses need to be docu-
mented. Documentation should not only
include determination of the etiology of
the illness and other clinical data but
should also include attempts to charac-
terize changes in lung function during
these illnesses, as well as when illness
free. This will necessitate the continued
effort in developing tests of lung function
which can be utilized in very young
children. Also, reference data for tests
of pulmonary function are needed which
have been collected in a longitudinal
fashion in illness-free, non-smoking
individuals 13-25 years of age who are
living in an area relatively free of
atmospheric pollution.
Table 2. P- Values for Tests of No Differences Between the Indicated
Regression Lines
Comparisons"
FVC FEV,
PEF FEF25-75%
Vmax75%
BF vs BM*
WF vs WM
BF vs WF
BM vs WM
Oc
0
0
0
0
0
0
0
<0.001
<0.003
<0.002
0
<0.001
<0.001
0
0
------- |