EPA 560/6-81-005
EPIDEMIOLOGY STUDIES
MAGNETIC LUNG MEASUREMENTS IN
RELATION TO OCCUPATIONAL EXPOSURE
TN ASBESTOS MINERS AND MILLERS OF QUEBEC
JANUARY 1981
FINAL REPORT
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF PESTICIDES AND
TOXIC SUBSTANCES
WASHINGTON, D.C.
-------
EPA 560/6-81-005
January 1981
MAGNETIC LUNG MEASUREMENTS IN
RELATION TO OCCUPATIONAL EXPOSURE
IN ASBESTOS MINERS AND MILLERS OF QUEBEC
by
David Cohen^,
Thomas Crowtherl
Maret Decklake -
Massachusetts Institute of Technology
Cambridge, Massachusetts
M
Project Officer
Jane E. Keller
Health and Environmental Review Division
Office of Pesticides and Toxic Substances
Washington, D.C. 20460
U.S. ENVIRONMENTAL
WASHINGTON,
PROTECTION AGENCY
D.C. 20*60
-------
DISCLAIMER
This project has been funded with federal funds from the
Environmental Protection Agency under contract No. 68-01-3859,
The content of this publication does not necessarily reflect
the views or policies of the U.S. Environmental Protection
Agency, nor does mention of trade names, commercial products,
or organizations imply endorsement by the U.S. Government.
-------
ACKNOWLEDGEMENTS
The authors wish to thank the Quebec asbestos Mining industry
and their employees for their cooperation, as well as the Institute
of Occupational and Environmental Health of the Quebec Mining
Association.
The authors are grateful to Mr. Ed Givler for his help during
the early phases of this project, to Mrs. Eleanor Pyle for her
encouragement and editorial help, and to Mrs. Donna Neuberg and the
MIT Statistics Center for their statistical evaluation of the data.
M. IK Don1clHlf.n in n n*v.^- Tnvrnl-i rj-iJ-nr nf tihP M^rl LcaJ— Re-SgHTTSK
i 1 of Canada*
Last we would also like to thank Emanuel Landau, Ph.D., Lenora
Barnes, M.P.H., and the entire staff at the American Public Health
Association; Joseph Seifter, M.D., Charles Poole, M.P.H., of the U.S.
Environmental Protection Agency for their cooperation and support.
The experimental measures and initial data analysis of this
project were supported by NIH grant #HL17543 and NSF grant #APR-761019,
-------
ABSTRACT
Fe 0. particles (magnetic) are often attached to asbestos fibers
(non-magnetic) in the primary asbestos industries,- therefore, a measurement
of Fe O could help determine the amount of asbestos in the lungs of
workers in these industries. As a first assessment of this method of
determining retained dust, magnetic measurements were made of the amount
of Fe,0. in the lungs of 115 miners and millers of chrysotile asbestos.
The performance of these measurements at an industrial site was found to
be feasible and practical. A relatively large amount of Fe O was seen
in the lungs of those with welding experience, which masked the Fe O
contributed by asbestos, therefore this group was considered separately.
For the remainder (non-welders), the amount of Fe O was plotted against
a total dust exposure index which was available for each individual. The
correlation between these quantities was not high, but was statistically
significant at the 0.01 level. For the non-smokers within that group, the
correlation was higher and the amount of Fe 0 was relatively greater.
These results suggest that the magnetic measurement of a chrysotile miner
and miller reflects, at least to some extent, the amount of asbestos in
his lung; the scatter could be due to individual differences in deposition
and clearance, to which this measurement should be sensitive. These
results are also consistent with the possibility that less dust is deposited
or retained in smokers than in non-smokers.
-------
INTRODUCTION
During the past decade, a method has been developed to measure
ferrimagnetic dust in the human lung (Cohen, 1973, 1975). In this method,
the lungs of the subject are first magnetized by an external magnetic
field. Then, after the external field is removed, the remanent field
produced by the magnetized particles is measured over the subject's torso;
this yields the amount of this dust in the lungs. Because chrysotile
asbestos often occurs with attached ferrimagnetic Fe O particles, and
preliminary measurements of several chrysotile miners and millers had
revealed a measurable amount of Fe 0 in their lungs, the question arose:
Can the magnetic method be useful in determining the amount of asbestos
in the lungs of miners and millers? The study reported here begins to
answer that question.
We concentrated on mining and milling because the Fe O content of
asbestos is high at this primary stage (Gibbs, 1971). In particular, we
concentrated on the miners and millers of chrysotile asbestos in
Quebec. For this well-studied population, previously the amount of asbestos
in a worker's lung had been inferred from his total dust index, defined
as (concentration of airborne dust) x (period of exposure), summed over
his various jobs (Gibbs and Lachance, 1972). Epidemiological studies
have shown a relationship between respiratory abnormality and this index
(Becklake et al., 1972; McDonald et al. , 1972; Rossiter et al. , 1972).
However, in some of these studies the correlations, while statistically
significant, have been low. This may be due to the indirect nature of
this index, which does not take into account individual variations
in dust deposition and clearance in the lung. The magnetic method, while
it has its own drawback of being completely dependent on the ratio of
Fe 0 to asbestos, may nevertheless be more direct. Our main objective
was to examine the relationship between the amount of Fe^ in the lungs
and the total dust index. If the amount of Fe 0 was indeed related to
the amount of asbestos in the lung, then we would expect at least some
correlation with this dust index.
The remanent field produced by the lung particles is very weak
-------
Cv> 10 gauss), in comparison to the background earth's field in which it
is measured (^ 1 gauss). This field must therefore be measured with care
and attention. While it can readily be measured in a well-equipped
laboratory, we previously had little experience in measuring this field
at a distant industrial site, such as at the mining town where this
study was performed. Hence our second objective was to see if such on-site
measurements are practical, not only for possible use in the primary
asbestos industries, but for other "magnetic" industries as well.
The measurements were made in 1974, and preliminary results had been
reported (Cohen, 1978). Following this study, other magnetic studies
of occupational groups were performed. These include two studies of
welders (Kalliomaki et al., 1978; Freedman et al., 1979) and a study of
coal miners (Freedman et al., 1980). In addition, a magnetic study of
stone workers is being completed (M. Kotani of Tokyo Denki University,
personal communication). However, in none of these studies was the
relationship examined between magnetic reading and dust exposure, nor was
the practical aspect evaluated. In addition to these studies, magnetic
studies of only small occupational groups have been described (Cohen, 1978).
MATERIALS AND METHODS
The Group Studied
The target group had already been selected for certain lung function
measurements (Peress et al., 1977), and were without radiologic abnormality
in that their most recent chest film was read as 0/0. This group contained
both smokers and non-smokers, an age range from 26-50, and a wide range of
total dust indices. These indices applied only to exposures up to 1967;
the period of 1967-74 was not included, but the omission is not likely to
be serious because the dust level had become reduced by the late 1960's.
Unfortunately, the general nature of the grouping changed after measurements
-------
began, when it was seen that workers with welding exposure had a relatively
large amount of Fe O in their lungs, which could mask the Fe O from
asbestos. We therefore divided the original group of 115 into those with
and without welding exposure. The relation between magnetic readings and
total exposure index was investigated only for the non-welding group, now
reduced to 51. For the remainder, our efforts were salvaged by estimating
their exposure to welding dust and examining the relationship between the
amount of Fe O in their lungs and this exposure; if the Fe O amount was
a reflection of occupational exposure, we would again expect to see a
correlation between the two.
The divisions within the group are given in Table I. Those called
smokers included 11 who were ex-smokers at the time of this study; they were
included because their smoking had taken place when most of their dust was
inhaled, some years earlier, hence it would have affected their response to
dust. Welders were divided into those with only a small (<0.1 year) and
those with a greater ( >0.1 year) exposure to welding dust, to allow various
correlations. However, it is seen that some correlations would be limited
by very small sub-groups, such as 5 or 12 welders.
Table I. Divisions Within the Group Studied
Smokers Non-smokers
Welders ( > 0.1 year)
Welders ( <0.1 year)
Non-welders
Total
33
5
33(33)
71
12
12
20(18)
44
Total
45
17
53(51)
115
*
Numbers in parentheses are those non-welders for whom total dust exposure
indices were available.
Magnetic Measurements
We measured not only the amount of Fe 0 in the lung, but also its
crude distribution within the lung; in addition, we measured two quantities
which are unique to magnetic particles. During the application of the
external field, the particles are both magnetized and become rotationally
-------
aligned with this field. However, their rotation is impeded by their viscous
environment; we measured a quantity involving this impedance, called the
apparent viscosity. After magnetization, the remanent field produced by
the particles is not steady but always decreases in time, typically by a
factor of three or four in an hour. This decrease, called relaxation, is
due to random rotations imparted to the initially-aligned particles by
local motions in the lung; because the remanent field is the vector sum of
the fields from all particles, it becomes reduced as the particles become
randomized. We measured the rate of relaxation. Our purpose in measuring
these auxiliary quantities was to see if they were useful in measurements
of occupational groups.
The magnetic measurements were made during each subject's visit,
mainly for his lung function tests, to a clinic in the town of Thetford
Mines. Between tests the subject made an initial 15-minute visit to the
magnetic station in the clinic, during which most of our measurements
were made. If he showed enough Fe,O., he then returned about 20 minutes
later for a 5-minute relaxation measurement; in some cases the schedule
allowed a third similar visit.
During the initial visit, the subject changed from his clothing
into special shorts, thereby removing all magnetic items from his person
(zipper, shoes, etc.); dental plates, which are always magnetic, were also
removed. In order to be magnetized, the subject was placed between two coils,
as shown in Fig. 1(A). These coils were powered by two car batteries, and
generated a magnetic field which was uniform over the lung to ± 10%. The
subject was magnetized twice. The first was only for viscosity purposes; a
field of 400-gauss strength was applied for only 0.35 seconds (called the
short pulse), which magnetized the particles but only partly rotated
them. After their remanent field was measured, the subject received the
second,main magnetization; a 750-gauss field was applied for 30 seconds
(the long pulse), which was enough strength and time to produce complete
alignment of particles. In Fig. 1(B), the remanent field due to the
magnetized particles is seen to be oriented almost horizontally, called
the z-direction, at the chest and back; the field is called B when
z
measured over those areas.
-------
Fig. 1. Sequence of magnetization and measurement. (A) Subject being
magnetized by the external field (broken lines); this field both magnetizes
and aligns the particles. (B) The remanent field around the torso produced
by the magnetized particles. Over the chest and back the field is
approximately horizontal, and is called BZ. (C) Subject performing a
measurement of Bz by moving up to the gradiometer, which is fixed to the
table. He is shown at the "near" position of his far-near-far motion. Not
shown here is a plastic shield which is mounted up from the floor, and
prevents him from touching the fluxgate and causing an artifact.
RF MF LF
10 SEC
Fig. 2. An example of the gradiometer output, due to measurements of a
subject's Bz at the lung points. The first three measurements are at the
subject's right front (RF) , middle front (MF) , and left front (LF) , recorded
about 30 sec after magnetization; the three back points (LB, MB, RB) were
recorded 15 sec later. The B ' s at R and L are larger than at M because the
£i
detector views more Fe3O4 there. The final three signals (back points
omitted) were recorded 27 min later; the typical decrease of Bz due to
relaxation is seen, as well as some baseline (background) disturbance.
-------
To measure B , we used the magnetic detector called the fluxgate.
This detector is compact, simple to use, and has a suitable sensitivity.
It can measure fields down to 5 x 10~ gauss in the 0-3 Hz bandpass in
which it was used, corresponding to the detection of about 0.05 mg of
Fe,O in the lungs. In dealing with the problem of the magnetic background,
the steady and fluctuating backgrounds are considered separately. To deal
with the steady background, the detector was always rigidly fixed in
position, and the subject moved up to the detector for a measurement; in
this way the change in detector output was only due to the subject's B .
z
The problem of the fluctuating background, due for example to moving cars,
(which act as large moving magnets), was minimized by using the fluxgate
in the gradiometer mode as follows. The output of the model we used
(#MF-5000, Automation Industries, Ltd.) is the sum of outputs from its
two identical probes; they were mounted in-line about 10 cm apart,
(horizontally, as shown in Fig. 1(C)), and oriented oppositely so that
the output was the difference in B between them. Thus a fluctuating B
z z
produced by a distant source was largely cancelled, while the B from a
z
subject's lung was not cancelled because it was much larger in the nearer
than the further probe. The fluctuating background in this mining town was
negligible when dealt with in this way.
For a measurement the subject first stood out of range of the detector,
then moved inward and placed a point of his torso at the detector as shown
in Fig. 1(C), then stepped back again. The detector has a bell-shaped
response curve in angle, with the maximum at the 0° or z-line and a
half-maximum at about ± 22°, corresponding to about ± 3 cm at the lung.
Measurements were made at three marked points on the chest and three
corresponding points on the back, called the six lung points; the chest points
were on a horizontal line 10 cm above the xiphoid, one at the midline and
two 10 cm on each side, at about the nipples. In addition, measurements
were made at the chin and three abdomen points, 20 cm below the chest
points. Measurements at these non-lung points were made in order to detect
any magnetic contamination occasionally present in the abdomen and the
head, which could mask the Fe30. in the lung. When such contamination was
seen, it could usually be demagnetized with a hand-held magnetic tape
eraser; in five subjects the contamination was large enough to resist this
procedure, and they were not included in the study.
-------
Processing the Magnetic Data
We here describe how the values of B from the six lung points were
Z
converted into the amount of Fe,O. in the lung and the auxiliary quantities.
The BZ values were first summed in these various ways: EF (three front B 's),
EB (three back B 's), EM (two middle B 's), EL (two left B 's), ER (two
z z z
right BZ'S), and E6 (all 6 B •s). By using the relaxation rates (see below),
these sums, from both short and long pulses, were extrapolated back to 0-time
(end of the magnetizing pulse). Next, the long-pulse sums were combined in
various ways to yield the crude distributions in the lung; EL/ER indicates
the amount of Fe O in the left lung compared to the right, 2EM/ (EL +ER)
indicates the amount in the middle compared to the average side, and EF/EB
is the front/back ratio.
To calculate the amount of Fe O , we first consider the simplest
relationship between this quantity and B ; this is when all the Fe O in the
Z J rr
lung is imagined to be compressed into a point source, called the magnetic
dipole; for which
B =2Imz~3 or B / (2 I m) = z~3 (1)
z z
where B is in gauss, m is the Fe,0. mass in grams, and z is the distance
Z J fi
to the field point in cm. I is a property of the magnetic dust called the
magnetization (in emu/gm); based on our measurements of laboratory samples,
and on literature values, we chose I = 10, so that eqn. (1) becomes
B /20m = z~ . For large z, an actual lung behaves as eqn. (1), with the z
z
falloff. However, for z < 20 cm, the B dependence on z is the curve shown
Z
in Fig. 3(A). To use this curve, if there were only one probe, say the near
probe of the gradiometer, we would simply use the z for that subject to
determine the ordinate B /20 m. and knowing B , we would solve for m. However,
z z
because of the far probe of the gradiometer (z+8.6 cm), we use the
difference of the ordinates, called F , so that
Fg = (1/20 m)
-------
where m is now the amount of Fe 0 in mg. The error involved in using this
formula is due to two sources. The first is a probable error of ± 30% in I;
since I is a constant for all subjects, this error does not affect their
relative m. The second is a probable error of ± 30% in F ; this is due to
uncertainty in individual lung spacing, hence this is the relative error in
m for the data presented here.
For the viscosity quantity, we defined and used this term:
apparent viscosity = Z6 / Z6 , which should increase with the
"^ J long short
viscosity experienced by the particles; however, it would also be sensitive
to the shape and size of particles. In laboratory measurements of this type
of quantity, we found it to be reproducible for an individual, but it does
not correlate with obvious variables, such as smoking or the residence time
of Fe O. in the lung.
Relaxation curves (B vs time) have been well studied; they have a
z
characteristic mathematical shape (Cohen, 1974, 1978) and a dropoff rate
which depends on the residence time of Fe O. in the lung and on smoking. The
shape of the curves is seen in Fig. 3(B); recently inhaled Fe_O. results in
a rapid dropoff such as the lowest curve of Fig. 3(B), while Fe_O. inhaled
years ago results in a slower dropoff, such as one of the upper curves.
Smokers who have recently inhaled Fe O show a much more rapid dropoff than
non-smokers (Cohen et al., 1979). It would seem, therefore, that the
relaxation rate of these workers might reflect both the residence time of
Fe,O. in the lung, and the amount of smoking. Because of time restraints,
which allowed only two or three B measurements following long-pulse
2
magnetization, the following system was chosen to quantify the relaxation.
Five relaxation groups were arbitrarily defined, .as shown in Fig. 3(B). Two
measurements >20 min apart were usually enough to place a subject in a
particular group; three or more measurements yielded increased accuracy
and allowed divisions within the group. Once the relaxation group was
determined, the various sums of B could be extrapolated back to 0-time.
The Indices of Exposure
Dust sampling had been carried out since 1948 at various locations
in the local mines and mills, using the midget impinger. This method yields
the density of respirable dust in the air, where the dust here consists of
both asbestos fibers and other particles. Using this density and the
8
-------
10
PURE DIPOLE (Z"3)
LUNGS
(EXTENDED
SOURCE)
10
15 20 25
Z-DISTANCE IN CM
10 20 30 40 50 60
MINUTES AFTER MAGNETIZATION
Fig. 3. Curves used in processing the magnetic measures. (A) Falloff from a
dipole source (z~ ) and from an extended lung. The extended lung falloff is an
average of falloff measurements from various subjects. The distance z is from
the lung center to the field point, estimated from the thickness of the
subject's chest. (B) The five relaxation groups which are used here. These
are bounded by four relaxation curves of characteristic shapes derived from
actual, measured curves; however, the four dropoff rates are arbitrarily chosen.
-------
month-by-month employment record of each worker, a total dust exposure index
was calculated, defined as (density) x (years worked at this location,
corrected to a 40-hour working week), summed over each job held by the
worker, up to the end of 1966. The index is given in units of (millions of
particles/cu ft) x (years), or mpy/cf. That the termination of the index
by the end of 1966 does not seriously affect its use for 1974 supported by
data on the steady decrease in dust density in the mills; for example, it had
fallen from an average of 75 mp/cf in 1948 to 10 mp/cf in 1966, with
reductions thereafter.
The exposure index for welding was derived more crudely, without dust
sampling. The index we used was simply the period of welding exposure in
years, corrected to a 40-hour working week; e.g., for a 10% work-time
exposure for 20 years, the exposure index was 2.0 years. A welding history
was obtained by careful questioning of each worker during the clinic visit
and, in cases of ambiguity, by telephone follow-up. In contrast to the
total dust index, the index used here applied up to the time of this study.
If a worker had any welding exposure whatever, for example if only to nearby
welding, he was grouped by us with the welders. Table I, therefore, shows
17 "welders" with exposure of only <0.1 year. We included torch cutting,
often used in the mines, as a welding procedure.
RESULTS
The amount of Fe,O , its crude distribution within the lung, the
apparent viscosity^ and the relaxation rate were readily measured; their
distributions within the group* are given in Figs. 4 and 5, and in Table II.
Fig. 4 shows that the amount of Fe3O4 in the lungs of welders is indeed
greater than in non-welders, as we had seen during the measurements. The
average amount of Fe 0. in the lungs is 1.3 mg for a non-welder, and
7.8 mg for a welder; if the first interval is excluded, the amounts increase
to 1.7 and 8.3 mg. There is therefore about fives times more Fe 0. in the
lungs of the welders of this group.
The results of one type of distribution measurement of Fe 0 within the
lung, the ratio of left/right, is shown in Fig. 5. The fact that less
* As distinct from Fe.,0. distribution within the lung.
'34
10
-------
20
CO
o
CO
en
UJ
a.
15
feio
oc.
UJ
CD
0
— NON-WELDERS
WELDERS
0.10 0.32 1.0 3.2 10 32
Fe304 IN MILLIGRAMS (LOG SCALE)
100
320
Distribution within the group of the amount of Fe 0 found in the lungs
A - - '
Fig. 4.
of the 53 non-welders, and the 62 welders. A log scale is~used, except for the
first interval which extends down to and includes zero. This interval is elevated
for the non-welders because it contains office workers, and for the welders
because it contains borderline cases where it was not certain that there had beer,
exposure to welding dust.
40 r
CO
•z.
o
CO
CC
UJ
Q.
30 -
-
0.4 0.
5 0.
8 1.
D 1.
1 1
2 1.4 1.6 1.8
O 20 h
cc
UJ
CD
10 -
0
u.^ u.o u.o
LEFT/RIGHT
Fig. 5. Distribution within the group of the ratio of the remanent field on the
left side of the torso to that on the right (ZL/ER). On the average, the left
side produces less field, hence contains less Fe O than the right side. This
is because of the reduced volume of lung on the left, due to the heart.
11
-------
dust is seen on the left, demonstrates that the magnetic method is
capable, by measurements at an industrial site, of yielding at least some
distribution information. The large left/right spread is due to our crude
B measurements; if the measurements would have been made with more
z
accurate positioning of each subject, then the spread would have been far
narrower. The remaining magnetic data are shown in Table II, along with
data on dust exposures. For the magnetic quantities, the total number of
subjects in each row (All) is limited to those who contain enough Fe 0 for
this type of measurement.
For the middle/side ratio, the table shows a most probable value of
about 0.5; this indicates the smaller amount of Fe O "seen" by the
detector at the middle, where there is almost no lung, in comparison to
the side. Again, the method reveals at least some distribution information.
However, the meaning of the most probable value of about 0.7 for the ratio of
front/back is not as obvious: although there is more posterior than anterior
lung, there are also differences in lung-to-detector spacing between front
and back, as suggested by one of the correlations presented below. The
large spread about the 0.7 value is again due to coarseness of the
z-spacing, to which this ratio is sensitive. In the next quantity, the
relaxation group, the coarseness has been reduced by averaging (16) , and the
spread about the most probable group II is actual Fe O behavior. This
applies to the apparent viscosity as well, where the most probable value
is about 1.8.
For the dust exposure data, the distribution within the group is
different for the years welding in comparison to the total dust index. While
they both have a large number in the first interval, due to "borderline"
welders for the former and office workers for the latter, the distribution
for the latter has an obvious dip in the middle interval. This was a
deliberate choice in the makeup of the group for the lung function
measurements; most workers were selected to have either low or high exposure,
but a small group was selected at the mid-level as well.
To visually examine the relationship between the amount of Fe_O and
the total dust exposure index, these quantities are plotted against each
other in Fig. 6 for all the non-welders, and in Fig. 7 for these
non-welders who are non-smokers. In both figures it is seen that there is
considerable scatter, with no obvious correlation between the two quantities;
12
-------
Table II. Number of Persons vs. Magnetic and Dust Quantities
Quantity
Middle/Side
Front/Back
Relaxation
Rate
Apparent
Viscosity
Welder's
Exposure
Total Dust
Exp. Index
Definition
2EM/(EL+ER)
number
EF/EB
number
group
number
E6 /E6
X. S
number
years
number
*
mpy/cf
number
0.0-
0
0.4-
7
< 1.0
0
<0.1
17
0-3.2
21
0.2-
4
0.6-
27
I
6
1.0-
7
0.1-
8
3.2-
8
Ir
0.4-
39
0.8-
26
II
40
1.4-
20
0.32-
13
10-
27
iterval;
0.6-
32
1.0-
16
III
15
1.8-
20
1.0-
18
32-
10
3
0.8-
20
1.2-
9
IV
2
2.2-
11
3.2-
2
100-
26
1.0-
2
1.4-
1
V
0
2.6-
10
10-
4
320-
20
1.2-
0
1.6-1.8
1
3.0-3.4
3
32-
0
1000-
0
All
97
87
63
72
62
112
Intervals are in log scales except first interval
this applies to both the points clumped near the origin, and the more visible
points. The same scatter and low visible correlation is seen in the welding
plot, in Fig. 8. In order to see if the correlations between these various
quantities were indeed as low as appears visually, a statistical analysis was
performed; the analysis was extended to include correlations between any of
the variables in this study where there might be some basis for correlation.
The analysis involved three computations. The first was least squares, which
provided the correlation coefficient r and a straight line; the second was of
the Spearman's rank correlation coefficient r (Snedecor and Cochran, 1967);
S
the third was a robust regression to provide a straight line, using an
iteratively weighted procedure in which the weights are inversely
proportional to the relative size of the residuals in the previous procedure
13
-------
10
8
CO
o:
CD
ro
ROBUST.
0
200 400 600
DUST EXPOSURE INDEX (MPY/CF)
800
Fig 6. Amount of Fe O in the lung of each worker vs his total dust
exposure index, for the 51 non-welders. Six points near the origin have been
omitted here. Although visual inspection shows no obvious correlation,
Spearman's rank correlation gives r =0.50 and the least-squares regression
s
gives r = 0.45; both are significant at the 1% level. The straight lines
are due to the least squares fitting of the data, and the robust regression
procedure.
14
-------
10
8
CD
S
< 6
cr
200 400 600
DUST EXPOSURE INDEX (MPY/CF)
Fig. 7. Amount of Fe 0 vs. total dust exposure index, for the 18 non-
welders non-smokers. Again visual inspection shows no obvious correlation,
but Spearman's rank correlation gives r =0.62 and least-squares regression
s
gives r = 0.80, again significant at the 1% level. The line slopes are
greater than in Fig. 6, therefore these non-smokers appear to have more dust
in their lungs than the smokers.
15
-------
5 10
YEARS OF WELDING
15
20
Fig. 8. Amount of Fe 0^ vs. years of exposure to welding dust, for the
45 asbestos workers with >0.1 years of welding. The arrow indicates that
one point is off the figure, at the upper right, with 29 years and 140 mg.
Again visual inspection shows no obvious correlation, but Spearman's rank
correlation gives rg = 0.37 (significant at about the 1% level), and
least-squares regression gives r = 0.88 (high significance). The least
squares line is pulled up by the high point.
16
-------
Table III.
Correlation Coefficients r due to Spearman's Rank Analysis
5
Group
Vital Capacity all
non-sm.
Chest
Thickness
Relaxation
Rate
Fe304
Density
low(104)t
low (39)
Fe304
Amount
low(106)
low (41)
Front/Back
all 0.38 (88)** low (76)
non-sm. low (28)
Middle/Side
all 0.21 (93)* low (76)
non-sm. low (28)
0.22 (93)*
low (32)
Appl. Viscosity all
non-sm.
low (64)
0.25 (67)*
0.49 (18)*
0.23 (72)*
0.47 (19)*
Relaxation Rate all low (76)
non-sm. low (28)
0.25 (72)*
0.40 (26)*
low (77)
low (27)
Cigarettes/Day all
low (77)
Years Welding
all
non-sm.
>0.1 yrs.
>0.1 yrs.
+ non-sm.
low (41)
0.64 (12)*
0.58(109)**
0.53 (40)**
0.55(115)**
0.50 (42)**
0.37 (45)**
low (12)
Total Dust Index all
(only non-weld.)
non-sm.
0.47 (50)** 0.50 (51)**
0.67 (18)** 0.62 (18)**
t r is simply called low (low correlation) if it is not significant at
the 5% level; the number of cases is given in parentheses.
* significant at the 5% level
** significant at the 1% level or less
17
-------
(Hosteller and Tukey, 1977)- This method was appropriate because it is not
sensitive to the many outlying points of our data to which the least squares
regression is quite sensitive.
The results of the Spearman's rank correlations are given in Table III;
the coefficient r is simply called "low" unless the correlation is
significant at the level of 5% or less, (indicating a probability of 5% or
less that this could have occurred at random). The chest thickness used
in the table had been coarsely measured, with the subject's back at the
wall. The relaxative rate had been quantified so that the number increased
linearly with the relaxation group, and subdivisions within a group were
also used. The density of Fe 0 was the amount of Fe O. divided by the lung
J 4r .5 ft
volume, obtained from their lung function measurements. The vital capacity
(% of predicted value) was similarly obtained; it can be an early indicator
of pulmonary abnormality due to exposure to asbestos dust, however the range
is small in this group because of their 0/0 chest films. For the cases of
direct interest (Figs. 6, 7 and 8), we here give least-squares results. For
the amount of Fe 0 vs. the total dust index, r = 0.45 for all non-welders
(51) and r = 0.80 for those that are non-smokers (18); these are both
significant at lower than the 1% level, in agreement with the rank
correlation. For the amount of Fe O. vs. the years of welding, r = 0.88
for those welders with >0.1 years exposure (45); this is certainly
significant at lower than the 1% level. Further comments on this table
are made below.
DISCUSSION
If the ratio of Fe 0 /asbestos was always a constant, both in respirable
dust and after deposition in the lung, then this study would not have been
necessary; a measurement of the amount of Fe_0. would directly give the amount
of asbestos in the lung. However, while this ratio is quite constant for the
airborne dust at most locations of the Quebec mines and mills, it is not
known if it changes in the lung. The ratio in more than 100 samples from the
mines and mills was recently measured (O. Djamgouz of Laurentian University,
personal communication); the ratio of Fe_0./chrysotile was always found to
be in the range of 2-5% by weight, where the lower the grade of asbestos, the
higher the percentage. In the lung, however, the Fe3°4 particles probably
-------
become detached from the fibers to which they were stuck, and may clear out
differently from the fibers; this is suggested by the fact that chrysotile
fibers in the lung are known to dissociate, after some months or sooner, into
fibrils (Suzuki and Churg, 1969). Further, there may be some inhaled
Fe3°4 dust which was not attached to fibers; and which may also clear out
differently. If the clearance rates are indeed different, then the ratio
of Fe3O4/asbestos will change in time; the greater the difference in
clearance rates and the variation in this difference from person to person,
then the greater the uncertainty in the magnetic method for determining
the amount of asbestos.
Assuming that the ratio is not constant, the straightforward way to
find out how well the magnetic method can determine the amount of asbestos
would have been to magnetically measure autopsied lungs of miners and
millers, and correlate the amount of Fe_O with the actual amount of
asbestos present. An alternative approach would have been to correlate the
amount of Fe O. with lung abnormality in miners and millers; here we would
have found out how well the magnetic method determines both the amount of
asbestos and its biological effects, which is another issue. Neither way
was available to us at that time, hence we chose a variation of the second
approach. This was to find the correlation between the amount of Fe O and
a dust exposure index which does correlate with lung abnormality; namely
the total dust exposure index. If the amount of Fe O. would correlate
highly with the amount of asbestos, then at best we can expect a low but
significant correlation between the amount of Fe O. and this total index:
this is because the correlation is not high between this index and lung
abnormality. On the other hand, if there was an intrinsically low
correlation between the Fe 0. and the asbestos, then we would expect the
correlation between the Fe 0. and the index to vanish.
Although the eye shows no correlation in Figs. 6 and 7, the statistical
analysis does show a correlation which is low but statistically significant.
This is compatible with a high correlation between the amount of Fe_O and
asbestos, and incompatible with a low correlation between them. Stated
otherwise, the magnetic method appears to indicate the amount of asbestos in
the lung, at least to some extent; the relative FeO /asbestos clearance
rates do not appear to vary greatly from person-to-person. It follows that
the large scatter in Figs. 6 and 7 must be due, in large part, to individual
19
-------
differences in deposition and clearance. The points well below the robust
lines, for example, are due to less deposition or more clearance than
normal (per unit of exposure), while the points well above the least squares
line are due to more deposition or less clearance than normal.
It is tempting to speculate that those points very high off the line,
say the two points near the upper left corner of Fig. 6, may represent
individuals at increased risk, in the sense that their dust clearance
mechanisms may be impaired. This would suggest that the magnetic method may
be useful in the primary industries for screening of workers with supposed
impaired clearance, hence at increased risk; although the dust levels are
now reduced, impaired clearance may nevertheless result in enough Fe3C>4 to
be measured. In that regard, we note in Table III that the amount or
density of Fe 0 does not correlate with the vital capacity, the possible
early indicator of abnormality; this strengthens the idea that the
measurement of amount of Fe.O. should be only for the purpose of determining
the amount of asbestos, not its biological effects; the increased risk
mentioned above would only be due to an increase of asbestos in the lung,
not its effects.
The conclusion that Fe O. indicates in part the amount of asbestos is indirectly
supported by the data on welders. Whether for all workers, as shown in
Table III, or for welders with >0.1 years exposure (Fig. 8), there is a
correlation between the Fe,O. and the welding exposure which is statistically
significant. Again, therefore, the magnetic method indicates/ to some
extent, the amount of occupational (welding) dust in the lung.
The large increase in the line slopes in Fig. 7 in comparison to Fig. 6
deserves comment. If statistically valid, this increase implies that
smokers have less dust in their lungs than non-smokers, per unit of
exposure. This would appear to be at variance with the,recent result (Cohen
etal., 1979) that smokers show impaired long-term clearance from the lung,
hence retain more dust than non-smokers. Because the dust-count (alveolar)
in that work did not begin until short-term clearance (from airways) had been
completed, this discrepancy could be resolved if one assumes that smokers
have far less alveolar deposition than non-smokers, perhaps because of
different airway diameters, etc. Stated otherwise, if this result is
valid, then in smokers perhaps less dust is deposited in the alveoli, but
this alveolar dust is cleared away more slowly than in non-smokers. But
20
-------
how believable is the slope increase?
On the one hand, Fig. 7 has only 18 points, with the large slopes
therefore depending on only a small number of special points. On the other
hand, the slope increase exists for both the least squares and the robust
lines. Perhaps we can only say that these results are consistent with
smokers having less dust in their lungs than non-smokers, for the same
exposure. Certainly we do not see here the large, reverse effect seen by
Cohen et al; in that case both slopes in Fig. 7 would have been smaller
by a factor of ten, which is clearly ruled out. It is unfortunate that the
group of non-smoking welders was too small (12) to see if their slopes would
also rise, compared to Fig. 8; Table III shows only a low correlation, hence
the line is meaningless.
Because we see that the magnetic method can be useful in the primary
asbestos industry for determining asbestos in the lung, we will comment
about various other aspects of the magnetic technique which have shown up
in this study, which may be of value in applying this method.
Concerning the practical aspects of these measurements, the fact that the
Fe_O was readily measured in a mining town does not insure that other
groups can easily be measured in other settings. It is a matter of
magnetic signal/noise; more Fe O. in the lungs produces larger gradiometer
signals which can override larger background noise. In this regard, the
average amount of Fe 0 in the lungs of different groups is important. We
saw that the average in this group was about 2 mg and 10 mg for non-welders.
We had measured some foundry workers while in the Thetford area and found
the average to be 140 mg(!)7 while measurements at the MIT lab of workers in the
secondary asbestos industries (finished products) yielded the much lower
average of about 0.10 mg. Foundry workers can therefore be measured in an
urban setting with high background; while workers in secondary asbestos
industries could only be measured with a detector more sensitive than a
fluxgate, in a magnetically quieter setting or with better
gradiometer-cancelling techniques.
We next consider the correlations in Table III of the auxiliary magnetic
quantities. There is a significant correlation between the front/back ratio
and chest thickness, indicating the front B 's are sensitive to the
z
thickness of the muscle and fat in the chest, therefore our methods of
21
-------
positioning subjects should be improved, perhaps with a standard E-spacing
used for all subjects. The same applies to the middle/side ratio, which also
shows some correlation; however, its correlation with the amount of Fe3C>4
is probably meaningless, because the correlation became lower with the
more uniform group of non-smokers. The applied viscosity shows correlation
with both Fe 0 amount and density, which is believable because there could
well be some sort of "clogging" viscosity with increased amounts of dust.
The relaxation shows a similar but lower correlation, again believable
for the same reason. However, the lack of correlation between relaxation
and amount of smoking is surprising, because this correlation in laboratory
measurements was high; one explanation could be that the latter result
was based on relaxation of dust recently inhaled, and perhaps the long
residence time of Fe O in our asbestos subjects had washed out the
correlation. Another surprising result with relaxation is not shown in
the table. The correlation between relaxation rate and the total dust
index (74 subjects including welders) showed r = 0.37, which is significant
s
at the 1% level. However, for the smokers within that group the correlation
vanished, so that the first correlation is not believable, and could be only
a random effect which is to be expected occasionally.
The data from Table III suggests, therefore, that the apparent viscosity
and relaxation rate are worth pursuing further as auxiliary quantities in
future measurements. The data also show that density of Fe O. in the lung
generally yields a somewhat higher correlation in comparison to the amount
of Fe3°4' this suggests that the density is a more useful quantity than the
amount of Fe,0 in the lung, and that this comparison should be pursued
further as well.
22
-------
REFERENCES
Becklake, M. R. , Fournier-Massey, G., Rossiter, C. E.,
McDonald, J. C. (1972). Lung Function in Chrysotile Asbestos
Mine and Mill Workers of Quebec. ARCH,.i ENVIRON. HEALTH 24,
401.
Cohen, D. (1973). Ferromagnetic Contamination in the Lungs and
Other Organs of the Human Body. SCIENCE 180, 745.
Cohen, D. (1975). Measurements of the Magnetic Fields Produced
by the Human Heart, Brain and Lungs. IEEE TRANS. MAG.-11,
#2, 694.
Cohen, D. (1978). "Report of the Low-Fiend Group: The Magnetic
Field of the Lung." Publication MIT/FBNML-78-1. National
Technical Information Service, Springfield, Virginia.
Cohen, D., Arai, S. F., and Brain, J. D. (1979). Smoking Impairs
Long-Term Dust Clearance from the Lung. SCIENCE 204, 514.
Freedman, A. P., Robinson, S. E., Goodman, L., et al (1979).
Non-Invasive Magnetopneumographic Determination of Luna Dust
Loads in Steel Arc Welders. CHEST 7_6, 352.
Freedman, A. P., Robinson, S. E., and Johnston, R. J. (1980).
Non-Invasive Magnetopneumographic Estimation of Lung Dust
Loads and Distribution in Bituminous Coal Workers. J.
OCCUP- MED. 22, 613.
Gibbs, G. W. (1971). Qualitative aspects of dust exposure in
the Quebec asbostos mining and milling industry. In
"Inhaled Particles III." (W. H. Walton, Ed.), Vol. II,
pp. 783-798, Unwin Brothers Ltd., Surrey, England.
23
-------
Gibbs, G. W., and Lachance, M. (1972). Dust Exposure in the
Chrysotile Asboestos Mines and Mills of Quebec. ARCH.
ENVIRON. HEALTH 2_4_, 189.
Kalliomaki, P. L., Korhonen, 0., Vaaranen, V-, et al (1976).
Lung Retention and Clearance of Shipyard Arc Welders. INT.
ARCH. OCCUP. ENVIRON. HEALTH 42, 83.
McDonald, J. C., Becklake, M. R., Fournier-Massey, G., and
Rossiter, C. E. (1972). Respiratory Symptoms in Chrysotile
Asbestos Mine and Mill Workers in Quebec. ARCH. ENVIRON.
HEALTH 24, 358
Mosteller, F. and Tukey, W. (1977). "Data Analysis and Regression."
Section 14H. Addison Wesley, Reading, Massachusetts.
Peress. L., Hoag, H., White, F., and Becklake, M. R. (1975). The
Relationship between Closing Volume, Smoking, and Asbestos
Dust Exposure. CLIN. RES. 23. 647A.
Rossiter, C. E., Bristol, L. J., Cartier, P. H., Gilson, J. G.,
Grainger, T. R., Sluis-Cremer, G. K., and McDonald, J. C.
(1972). Radiographic Changes in Chrysotile Asbestos Mine
and Mill Workers in Quebec. ARCH. ENVIRON. HEALTH 24, 388.
Snedecor, G. W., and Cochran, W. G. (1967). "Statistical Methods."
6th Edition, p. 194. Iowa State University Press, Ames, Iowa.
Suzuki, Y., and Churg, J. (1969). Structure and Development of the
Asbestos Body. AM. J. PATH. 55, 79.
24
-------
APPENDIX
The preceeding text, without this appendix, has been submitted
for journal publication. The material in this appendix therefore
is available only in this report. The material consists of
details of methods (Appendix A) and of the results (Appendix B),
and is intended for those who wish more knowledge of this work.
However, Appendix A contains some details which are useful for
magnetic measurements of the lung generally. The style of presen-
tation in the appendix is occasionally more casual than in the
main text.
25
-------
APPENDIX A. SOME DETAILS OF METHODS
Performing the Magnetic Measurements
We first give a brief outline of the procedure we used during the
first, 15-min visit of the subject to the magnetic station. We then pre-
sent some details involved in this procedure.
Outline
1. The test is explained to the subject; the subject gives his I.D. and
welding (torch cutting) history. Next the subject removes all his
clothing except socks, and changes into athletic shorts furnished by
us.
2. The chest thickness is measured, as well as the subject's chest height.
Three points are marked on his chest on a horizontal line 10 cm above
the ziphoid, and 10 cm apart. The height of the platform between coils
is adjusted so that the subject's chest will be centered between the
coils.
3. The subject is magnetized with a short pulse of 0.3 sec, of strength of
about 400 gauss. Exact pulse length is noted on an oscilloscope.
4. Subject's chin and abdomen are checked for magnetic interferences by
measurement with the flux gate gradiometer.
5. Guided by the operator, the subject undergoes his far-near-far motions,
touching the shield for each "near" position. First the three chest
points are measured, then the three back points.
6. The subject is magnetized for 30 sec at 750 gauss.
7. Again the gradiometer measurements are made for each of the chest and
back points.
8. Step 7 is repeated approximately 20 min later, then again later if the
subject's time permits.
Various Details
The MF-5000 gradiometer was arranged so the outer tips of the probes
26
-------
were 4 11/16 inches apart (hot-spots were 3 3/8 inches apart). The MF-5000
was always used on the 1 milligauss range; it fed a Varian recorder used
on either a 10, 100, or 1,000 millivolt range. The chart speed was always
4 inches per minute. A calibrated coil was always present on the sensor
closest to the subject, fed by the current from a calibration unit. It
yielded either 1, 5, or 10, or 50 or 100 or 500 or 1,000 or 5,000 or
10,000 times 10~ gauss, applied to one probe only. The capacity across
the output of the MF-5000, which filtered out the higher frequencies, was
850 mf; this yielded a 3db point somewhere between 0.5 and 1.0 Hz, with a
fall-off of 6db per octave, since filtering took place only at one point
in the circuit; the chart recorder also had its own, slow time constant.
The two large magnetizing coils were set with an inside spacing of
12.0 inches, and were connected in series and fed by two storage batteries
in series (24 v), yielding 750 gauss. When the switch was first closed, a
current of about 400 amps flowed, which drooped during the 30-second appli-
cation time to about 375 amps. During the 30-second magnetization, each
subject was rocked from side to side slowly, in order to get fairly uni-
form magnetization across the entire chest; he was not moved up and down,
however, or in and out, but placed about half-way, fore and aft. One sub-
ject was too heavy to fit between the magnets, and he was magnetized in-
stead with a powerful Alnico magnet (about 250 gauss at the lung), slowly
rubbed over his front and back.
Various difficulties were encountered during the measurements of
the first few subjects; these included inability to exactly position the
measuring points at the fluxgate, much magnetic contamination on the skin
especially around the chin and neck, and the phenomenon we later called
spiking. To cope with the first difficulty, we experimented with the
subject's posture. During the first day or two, we asked the subject to
"hang loose" while placing marks and making measurements. Subsequently,
we found that this led to the trouble of back measurements, which were
too low on the back; it was better to have the subject stand erect at all
times, so that the back measurement lined up more on the upper part of the
back, closer to the center of activity. Concerning the next problem, we
began to vacuum the chest and often the back, certainly the neck, of each
27
-------
subject. We also tacked Kleenex to the subject's chest and back, in order
to prevent magnetite in the air from clinging to his skin during magnetization.
For spiking, we eventually found, by trial and error, that if we covered
the front of the lucite shield with ordinary brown masking tape, the spiking
disappeared. The spiking therefore is probably associated with some
electrostatic discharge, at frequencies to which the fluxgate is sensitive.
That is, skin rubbing against the lucite appears to generate charges in
some fashion or other; this does not appear to be the case with skin against
masking tape.
One phenomenon we noticed, as yet unexplained, was with subjects
who had very high readings, such as some welders. It was noticed, when
they stood at the "far" position, that we could see their breathing on the
chart recorder. The amplitude of this modulation seemed to be far greater
than one would expect at that distance. We should also note that, when a
subject stood at the far position, with perhaps 10 or 12 inches between his
skin and the lucite, the chart reading was not a true zero B ; when a subject
was asked to do a 180° turn at this far position, and we could see the
modulation of the baseline accordingly, all readings should really be corrected
upward by a certain percentage, because of the distance effect; that is,
subjects were not adequately far away in their far positions. It appeared,
on visual inspection, to be a correction of perhaps 3 or 5%. Another point,
of some interest, is the modulation of the signal during breathing, by
breathing, when the subject is at the lucite. It has been noticed that when
the subject stands with his front pressing the lucite, there is almost no
modulation by breathing; however, when his back is against the lucite, it
is almost always seen that there is a large modulation of the signal.
For investigators using this technique for determining magnetic
particles in the lung, the following list of interferences (both correctable
and non-correctable) may be useful.
It is, of course, important that non-magnetic clothing be worn, and
that shoes, belts, watches and false teeth be removed. In this study white
athletic shorts with elastic waist bands were provided, although hospital
trousers with drawstrings would have been preferable. Magnetic checks at
neck height revealed steel pins in teeth; they also revealed a magnetic
piece in a toupe! Abdominal checks showed up steel sutures from surgery,
28
-------
and often steel chips ingested with food. Steel items in the mouth and
abdomen were often correctable by using a large but hand-held magnetic 60-Hz
eraser. Steel particles embedded in the skin were occasionally erasable
by a short-range eraser.
The sensitivity of the fluxgate gradiometer was limited by its rms
noise of about ^5x10 gauss (rms //Hz"). In addition to this intrinsic
noise the field changes produced by passing cars and trucks would occasionally
introduce a total reading uncertainty of up to 10x10 gauss. Assuming that
they were random, this corresponds to an uncertainty in calculated total
Fe3O4 from one reading of ±0.3 mg of Fe 0 or ±0.1 mg for the average of
12 readings, because
a= a//n
where a is the standard error between the sample mean and population mean,
a is the standard deviation of a single measurement,
n is the number of trials (12 here).
For subjects having a magnetic lung burden so great that it exceeded
the background noise by more than two orders of magnitude, it was found that
difference in any two consecutive readings of the same point averaged about
9%. This variability is a function of how accurately the subject aligned the
point on his body with the detector and on how hard he pressed against the
plastic shield. These uncertainties are negligible compared to the ± 20%
uncertainty which we believe was introduced by using an average chest thickness
correction curve (Fig. 3 (A) ) ; they are also negligible compared to the ± 30%
uncertainty in the I value , which depends on the particle size and shape of
RS
the Fe O in the lung. However, as noted in the text, this latter error is
an absolute error, similar for all subjects, and does not affect the relative
amount of Fe O. between subjects.
Processing the Magnetic Data
Obtaining the Amount of Fe O in the Lung from Measurements of the Remanent
Field
In order to convert gradiometer readings into amount of magnetite in
the lungs, two factors must be used. The first is the remanent magnetic
moment of the Fe_O. particles in the lung following magnetization in the
coils. It will be shown that this value depends on the grain size, aspect
29
-------
ratio, and orientation of the magnetic particles. The second is a factor
to account for the distance between the lungs and the detector (which
varies with the individual subject's chest thickness, lung thickness and
shape) and for the magnetic particle extended distribution within the
lungs.
1. The Magnetic Moment
The remanent magnetization of dispersed magnetite powders has been
measured by Parry (Parry, 1965) who obtained good agreement compared to the
theory (Stacey and Banerjee, 1974). The theory uses
IBS-VN (1)
where I is the remanent magnetization in emu/cc,
Ro
H is the coercive force in oersteds, and
c
N is the effective self-demagnetizing factor for the grains
(For spherical grains N=4ir/3).
Experimentally Parry found that for a 1% dispersion
I_ = 3.0xlO~3H (2)
JxO C
which means (using a factor of 100) that the value of N is 3.32, i.e., 0.79
times that of a spherical grain. This means that some of the grains were
elongated and aligned in the direction of the applied field.
Another useful relationship is that H «
-------
5 mg into 2 cc of an epoxy (EZ Mount plastic, Donsel Equipment Co., Westboro,
MA) in a clear plastic cubical box 2 cm on a side. Pairs of samples differed
by the magnetic field environment in which they were placed while the epoxy
hardened; unoriented samples were placed in a magnetic shield, while oriented
samples were placed in a field of ^ 700 gauss.
We then measured the remanent field of the samples using a fluxgate
probe (MF-500, manufactured by Automation Industries) in the MIT shielded
room. A single probe was used in these experiments (the second probe of
this gradiometer was placed in a separate shield).
The procedure before measuring each sample was to first demagnetize
it in a 60-Hz demagnetizer, then magnetize it in a 750-gauss field for 30
seconds. The remanent magnetization was calculated from the formula for the
field from a magnetic dipole
2 Ips V 2 I m
B = rc = RS (4)
z3 z3p
where m is the mass of magnetite in grams,
I is the remanent magnetization in emu/cc (and is therefore of
Ro
of different dimensions than the I used in the text),
z is the distance between the center of the dipole and the
detector in cm,
p is the density of Fe O in grams/cc,
B is the axial magnetic field in gauss.
Z
Typical values for an unoriented sample were m=4.8 mg, z = 12. 7 cm, p = 5.2
"3 —6
am/cm. B =35x10 gauss; this gives I =38.9 emu/cc or 7.5 emu/gm. Since
z R*-3
the saturation magnetization of magnetite, I , is 450 emu/cc, I /I =0.086.
t> Ro S
An independent calculation of I / I was obtained by measuring the
i\O O
coercive force H .for an unoriented sample in a PAR Vibrating Sample Magnetometer
c
Model CC-1. Here an H of 150 Oe was measured. Using eqn. (1) with N=4ir/3,
this gives I /I =0.080, in good agreement with the direct measurement.
Ro o
For the samples which were oriented in the 700-gauss field while the
plastic hardened, the measured value of 1^ was 64.5 emu/cc, or 12.4 emu/gm,
*No longer manufactured by them. The basic design is by the Forster Co. of
West Germany, and their units are available.
31
-------
which is 70% greater than for the unoriented particles, or I /I =0.143.
I\O O
The increase is due to the decreased demagnetizing factor N for particles
with their longer axes aligned in the direction of the applied field. Since
it is assumed that at least some of the Fe,O particles in the lungs rotate
during the time the field is applied, a value of I of 10 emu/gm (or 52 emu/cc)
rd>
is used in this study, since it lies between the unoriented and completely
aligned values.
2. The Spatial Factor: The Extended Lung vs. the Dipole
For the actual extended lung, we can assume that its z-dependence can
be approximated as that magnetic dipole (z ) for distances greater than
20 cm between the center of the lung and the detector. For distances less
than 20 cm it is found that the B for the human lung falls off less rapidly
_3 z
than z . This falloff distance factor, F (z) , is plotted as the lower curve
m
in Fig. 3 (A) ; (the sub-m refers to a magnetometer or single-probe measurement,
as opposed to a gradiometer, the latter designated as sub-g; we also note that
in the tables F (z) is used as F (d) , as they are meant to be interchangeable
g g
here). This curve is an average of two experimentally determined curves,
one for a tall and thin subject, the other for a short and stocky subject.
F (z) is defined as
m
F(z) -
m ' r
where M is the total remanent magnetic moment.
The net reading for the gradiometer used in this study is the difference
between two probes whose sensing elements were 8.6 cm apart. Thus the
gradiometer distance factor, F (z) , is defined in terms of the magnetometer
distance factors, F (z) as
m
F (z) = F (z) - F (z + 8.6) , and (6)
g mm
Bg(z) = 2MrFg(z) = 2IRsmFg(z) /p (7)
where B (z) is the average gradiometer field reading in units of 10 gauss, or
B (z) =1/6 E 6 (8)
g t=0
where Z_ 6 is the sum of the six gradiometer readings taken at the six lung
points; these are three on the chest, three on the back, each extrapolated
back to time zero, which is just at the end of the magnetization in the
coils. Combining egns. (7) and (8) and solving for m gives
32
-------
m (mg of Fe O.) = [83.3 p E 6j 10 3 / [F (z) I 1 (9)
•3 1 t_0 g RS
and since I = 52 emu/cc and p = 5.2 gm/cc, then
m (mg of Fe O ) = [8.33 E 6] 10~3 /F (Z) (10)
34 t=Q g
or since the average lung volume, V., is equal to FRC+ TV/2,
the density m/V =[8.33 E 6]lO~3 / {F (z) [FRC + TV/2] }. (11)
*• t=o g
References
Dunlop, D. J. (1972). Magnetite: Behavior near the Single Domain Threshold.
Science 176, 41.
Parry, L. G. (1965). Magnetic Properties of Dispersed Magnetite Powders.
Phil. Mag. 1.1, 303.
Stacey, F. D., and Banerjee, S. K. (1974). "The Physical Principles of Rock
Magnetism." Elsevier Scientific Pub. Co., New York.
33
-------
APPENDIX B. SOME DETAILS OF RESULTS
In this part of the Appendix two types of details are presented of
the results of measurements. The first type consists of data of each of
the 115 individuals. The second type consists of bar graphs which illus-
trate the distribution of quantities within the group. The individual
data is given in the following Table BI. We here explain the column head-
ings of that table.
SUBJECT NO.: This is the subject number, only for purposes of this table.
We are following the rule of preserving the subject's anonymity.
AGE: This was the subject's age at the time of the study, in 1974. For
purposes of the lung function tests for which this group was chosen,
the ages were confined to the ranges 25-35 and 41-50. In addition,
three individuals were measured magnetically whose ages are not shown
(#3, 5, 97), but who are over 50; there were no dust indices
and other data available for them, hence they were not included in the
main correlations.
HEIGHT: Included only as information on the body build of the subject.
WEIGHT: Included only as information on the body build of the subject.
CHEST TH: This is the chest thickness, used to calculate the distance from
the lung center to the nearer probe, for application to Fig. 3(A) of
the text. The distance used (d) was (half the chest thickness) +1.5 cm.
Thickness is here given in inches (not uncommon in 1974J)
LUNG VOL(£): This is the lung volume (in liters), used for calculating the
density of Fe 0 in the lung (further on). We calculated this volume by
multiplying the 'FRC (measured by the McGill Group during the clinic
visit) by 1.08, which is an approximate way to include 0.5 of the tidal
volume.
V.CAP: The vital capacity, also measured by the McGill group during the
clinic visit, expressed as % of predicted value; because the subjects
were chosen to be without pulmonary abnormality, the range of values is
restricted accordingly.
34
-------
CIGS/DAY: The number of cigarettes smoked per day, up to the time of the
study. Ex-smokers are indicated by ( ); in our correlations, we included
these with the smokers. The values in this column were determined by care-
ful questioning by the McGill group.
YRS WELD.: The years of welding exposure, as defined in the text (METHODS).
The designation < 0.2 indicates that, as far as we can determine, the
subject had some welding exposure, but certainly less than 0.2 years.
A bar under a value indicates that the welding exposure was recent, as
opposed to "old"; recent was defined as greater than 0.10 of his peak
yearly exposure was obtained during the past year (ending in Aug. '74).
DUST: This is the total dust exposure index in units of (millions of par-
ticles/cu ft) x yrs. As explained in the text, it is based on midget
impinger counts, etc., and only applied up to the end of 1966. There
was no inclusion for dust exposure after that; however, for the period
not included, the dust levels were generally much lower than previously.
I6(10~'g): This is the sum of the BZ values measured over the six lung
locations on the torso; it is 6 x (the average remanent field from the
lung). It is used in the formula (see METHODS) to yield the amount of
Fe 0 . A designation such as 40->75 indicates that the "true value"
can be anywhere in that range, i.e., the error is greater than normal.
CONV.(F (d)): This is the geometric conversion factor involving both the
extended lung and the gradiometer (differences), as explained in METHODS
(both in the text and in App. A).
AMT.Fe-jO-: After the data was processed and prepared, it was seen that
(REVISED)
F (d) should have been revised downward by about 13%, with
variations among individuals; hence, the amt. Fe 0 should have been
revised upward by a corresponding amount. This new, revised value was
not used as data in the text. For future use, it is a better value than
the unrevised value.
AMT.Fe-aO : This is the unrevised value, actually used in the text for the
(UNREV.) measured amount of Fe 0 in the lungs of a worker.
DENS.: This is the density of Fe 0 in the lung, calculated by dividing
the revised amount of Fe 0 by the lung volume.
35
-------
MIDDLE/SIDE: This is information re distribution of Fe 0 in the lung. It
is the middle B , summed front and back, divided by the sum (front and
z
back) of the left side, and the right side. It is otherwise written as
2ZM/ZL+ZR.
LEFT/RIGHT: This is also the distribution within the lung and is the two
left points divided by the two right points, otherwise written as ZL/ZR.
APP.VISC.: This is the apparent viscosity, defined as Z6 from the long
pulse, divided by Z6 from the short pulse. Numbers with ( ) indicate
larger errors than usual.
RELAXATION: This is relaxation group into which this subject fits. A
minus after the group refers to the lower range within that group, a
plus refers to the higher range within that group, and no sign refers
to the middle of the group. It is a finer designation than just the
course grouping given in the text RESULTS. The ( ) refers to a larger
error than usual, perhaps by a half-group; e.g. (Ill) could mean III-
or III+.
36
-------
TABLE BI. DATA OF EACH SUBJECT
co
SUBJECT NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
W
a
49
49
-
49
-
28
43
49
44
50
43
50
47
49
47
48
50
50
46
49
31
48
28
HEIGHT (cm)
160
170
173
155
167
174
173
162
172
168
167
174
175
166
176
162
167
172
167
161
165
167
178
WEIGHT (kg)
71
75
71
57
64
63
71
64
83
71
75
89
75
90
88
57
61
70
60
58
76
62
84
"c
•H
E^
EH
U)
W
s
10.4
10.6
9.8
9.2
9.5
9.4
9.6
9.6
10.9
9.8
9.6
11.2
9.8
12.0
10.0
9.0
8.8
9.8
8.9
9.2
10.0
9.0
9.8
<••*
=><
I
O
,J
1.8
4.1
-
2.1
-
5.0
2.7
2.4
4.1
2.9
-
2.6
3.8
1.7
2.9
3.2
2.7
2.9
2.9
3.7
2.2
3.5
2.9
'M
a
Of
1
>
87
94
-
78
-
116
-
88
130
114
104
111
112
87
93
120
96
92
114
-
112
97
119
CIGS/DAY
0
20
20
20
20
0
20
50
(20)
0
0
20
0
(8)
(25)
22
20
20
0
0
0
22
0
•
3
£
2
>H
<0.1
<0.1_
0
0
0
<0.1
1-2.
0
0
<0._1
<0.2^
0
0
0
0
0
4. £
0
0
2.2_
0
3-2.
<-o.i
DUST (mpy/cf)
26
107
—
589
—
1
116
5
10
54
45
111
588
665
407
618
525
106
29
30
2
692
40
"5»
t^
o
•H
10
W
0->€5
665
1140
590
380
360
5610
40+75
480
2050
1450
0->-70
2710
285
125
145
2800
265
550
1300
145
1210
335
*••*
3
*"&>
&4_
8
1.63
1.56
1.80
1.97
1.88
1.93
1.85
1.85
1.50
1.80
1.85
1.43
1.81
1.23
1.74
2.05
2.10
1.80
2.07
1.97
1.74
2.05
1.80
AMT.Fe304(ing
(REVISED)
<0.4
3.6
5.3
2.5
1.7
1.6
25.3
0.2
2.7
9.5
6.5
<0.5
12.5
1.9
0.6
0.6
11.1
1.2
2.2
5.5
0.7
4.0
1.6
£
^r
o -~
ro .
•
OH
PH
<
._
2.8
1.8
2.2
1.9
_
1.3
__
2.6
_
2.3
_
1.9
2.4
(2.9)
1.8
1.8
1.55
-
2.6
-
1.6
2.5
RELAXATION
II-
11+
II
11+
11+
II
II-
1+
II
III
II
(III)
(1+)
II
(ID
II-
III-
(IV)
III-
(I)
-------
TABLE BI. DATA OF EACH SUBJECT (CON'T.)
00
SUBJECT NO.
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
W
0
50
42
47
31
43
47
45
34
49
28
31
48
45
44
26
47
43
27
31
48
46
48
41
HEIGHT (cm)
163
164
179
173
178
165
173
171
187
171
189
186
174
167
181
163
179
169
172
177
177
161
172
WEIGHT (kg)
86
68
96
82
66
60
80
70
80
77
82
92
74
67
75
66
74
88
80
84
73
87
82
'H
•H
EH
EH
CO
W
5
9.8
9.6
11.4
10.0
9.0
9.0
10.2
10.2
10.6
9.4
9.5
11.4
10.5
9.2
10.0
10.5
9.2
10.5
10.5
11.0
9.4
11.2
10.6
o
2.1
2.4
1.6
3.2
4.1
2.9
2.7
3.0
5.2
3.7
4.0
3.2
2.6
4.4
4.5
2.7
3.8
3.0
3.0
2.4
2.8
2.8
2.7
a
df
I
83
104
110
109
103
-
116
98
89
120
112
133
98
106
109
92
109
115
114
101
102
118
91
CIGS/DAY
0
(10)
0
15
15
18
12
15
18
0
0
35
25
0
(10)
18
9
0
15
0
0
40
30
a
s
co
K
0
0
< 0.1
0
3.0
1.2
2.6_
2-1
0.5_
1.0_
<0.1
0
0
0
<0.1_
0.8
18-2
0
0._3
0.4
18.0
0.2
<0.1
DUST(mpy/cf)
187
9
442
1
123
117
124
20
17
1
3
22
125
214
23
579
10
2
21
507
73
133
18
0
rH
tO
0+65
0+70
8430
0+70
630
675
800
2570
1310
2140
85+95
0+105
140
350
1190
1250
12300
0+70
425
240
7020
460
35+80
•a
tn
8
1.80
1.85
1.37
1.74
2.05
2.05
1.66
1.66
1.56
1.93
1.88
1.37
1.60
1.97
1.74
1.60
1.97
1.60
1.60
1.46
1.93
1.43
1.56
AMT.F6304 (mg
(REVISED)
<0.2
<0.2
51.3
<0.4
2.6
2.7
4.0
12.9
7.0
9.2
0.4
<0.7
0.7
1.5
5.7
6.5
52.0
<0.4
2.2
1.4
30.3
2.7
0.2
o —
n •
0) >
<0.2
<0.2
45.0
<0.3
2.0
2.1
3.3
11.0
5.9
7.3
0.3
<0.6
0.6
1.2
4.6
5.4
40.0
<0.3
1.8
1.2
24.0
2.3
0.3
DENS. (ygm/cc
MIDDLE/SIDE
0.0 -
0.0 -
2.0 0.4
0.0 -
0.6 0.3
0.9 0.2
1.5 0.2
4.3 0.4
1.3 0.3
2.5 0.4
0.1 0.1
0.0 -
0.3 0.4
0.3 0.3
1.3 '0.3
2.4 0.4
14.0 0.2
0.0 -
0.7 0.5
0.6 0.3
11.0 0.2
1.0 0.3
0.8 -
FRONT/BACK
-
-
0.7
-
1.3
1.3
0.8
0.7
1.0
0.9
-
-
0.8
0.9
0.8
0.7
0.8
-
1.2
1.1
1.2
0.7
-
LEFT/RIGHT
-
-
0.9
-
1.0
0.7
1.2
0.7
1.1
1.0
1.0
-
0.9
0.6
1.0
1.4
0.7
-
0.8
0.6
1.1
0.8
1.0
u
co
H
1
-
-
2.9
-
1.5
-
1.9
2.0
2.0
2.5
(1.5)
-
-
1.9
1.7
2.8
1.3
-
1.3
(2.2)
2.6
1.4
-
RELAXATION
-
-
Ill
-
II
II-
III-
I
II-
11+
-
-
-
II
11+
III-
III+
-
II-
(III)
nn-
iit
-
-------
TABLE BI. DATA OF EACH SUBJECT (CON'T.)
w
SUBJECT NO.
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
w
u
<
41
44
50
50
49
46
47
44
50
27
46
45
43
43
46
43
46
49
43
46
44
48
50
HEIGHT (cm)
172
175
173
174
155
173
166
175
175
180
178
167
172
171
168
172
170
176
178
162
159
163
177
WEIGHT (kg)
70
73
89
89
49
75
60
78
77
94
74
70
67
77
66
73
65
61
63
79
64
89
87
*-x
e
•rH
EH
EH
CO
W
5
9.6
10.2
10.5
11.4
8.0
9.5
9.5
10.5
11.2
10.8
9.0
11.2
9.6
10.2
9.0
9.5
9.0
9.5
8.2
10.0
10.0
12.0
11.2
^^
o*
I
CJ
^
3.8
4.8
1.9
-
2.2
2.8
3.3
3.5
3.2
3.0
3.2
2.5
3.6
2.6
2.6
2.4
2.9
4.1
3.9
2.5
2.6
3.0
2.7
V.CAP(%pr)
127
95
115
-
51
103
127
80
113
104
85
101
144
105
87
99
88
96
94
106
98
-
80
CIGS/DAY
15
25
0
(50)
(12)
20
15
0
0
0
20
0
0
(50)
38
20
0
25
55
0
(25)
12
25
•
Q
^
3
>•
15. £
<0.1
2-i
0
0
0
0
0
0
1.3
0
0
0
<0.1
0
1.2
0
0
0
1.5_
0
0.8
0.8_
DUST (mpy/cf
11
9
30
835
534
21
11
227
72
32
116
7
7
57
110
13
154
641
8
523
308
119
644
f~*.
Cn
r^
0
•H
10
W
9530
170
455
1160
525
560
0+70
570
450
630
130
35+65
0+70
150
285
150
230
1140
245
2900
1400
430
1200
^^*
s
r01
CM
•
O
U
1.85
1.66
1.60
1.37
2.35
1.88
1.88
1.60
1.43
1.51
2.05
1.43
1.85
1.66
2.05
1.88
2.05
1.88
2.30
1.74
1.74
1.23
1.43
AMT.Fe304(ir
(REVISED)
42.9
0.8
2.4
7.0
1.9
2.5
<0.2
3.0
2.6
3.5
0.5
0.2
<0.2
0.8
1.2
0.7
0.9
5.0
0.9
13.9
6.7
2.9
7.0
>:r
0
n ^
0) .
frj £>
s* B
34.0
0.7
2.0
6.2
1.3
2.0
<0.2
2.5
2.3
2.9
0.4
0.2
<0.2
0.6
0.9
0.5
0.7
4.0
0.6
11.0
5.4
2.6
6.1
o
1
•
w
z
3
11.0
0.2
1.3
-
0.9
0.9
0.0
0.9
0.8
1.2
0.2
0.1
0.0
0.3
0.5
0.3
0.3
1.2
0.2
5.6
2.6
1.0
2.6
MIDDLE/SIDE
0.2
0.3
0.5
0.3
0.3
0.5
-
0.4
0.4
0.3
0.2
-
-
0.5
0.4
0.4
0.4
0.5
0.2
0.3
0.4
0.3
0.3
FRONT/BACK
1.4
1.1
1.1
-
1.1
0.8
-
0.8
0.9
0.8
1.2
-
-
0.9
1.1
0.9
0.8
1.1
1.0
1.1
0.8
0.6
0.6
LEFT/RIGHT
1.0
0.7
1.1
1.1
0.6
1.1
-
0.8
0.8
0.9
0.9
-
-
1.2
1.1
1.0
0.9
1.0
1.4
1.6
0.8
0.9
0.7
•
O
C/3
H
>
•
ft
8!
2.8
(1.5)
-
1.6
2.1
2.4
-
-
2.0
2.3
-
-
-
(1.6)
1.1
-
-
2.7
(4.0)
2.1
-
(2.7)
2.6
RELAXATION
II
-
(Ill)
III
III
(II)
-
II
II
1+
-
-
-
-
Ill
-
-
II
1+
III
11+
11+
11+
-------
TABLE BI. DATA OF EACH SUBJECT (CON'T.)
SUBJECT NO.
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
w
34
48
43
28
48
45
42
31
46
49
41
49
29
35
43
27
45
26
32
27
41
31
46
HEIGHT (cm)
165
170
181
176
168
164
174
178
161
,161
175
175
178
173
172
164
166
172
169
176
164
176
175
WEIGHT (kg)
53
61
75
77
66
71
69
60
89
73
69
63
72
75
66
75
85
65
56
56
54
113
84
•H
CO
W
5
8.0
10.0
9.2
10.0
9.5
9.0
9.4
8.4
10.9
11.0
10.2
9.4
9.0
10.6
10.2
10.0
11.5
9.0
8.0
8.0
8.4
12.0
11.0
1
2.9
3.6
3.7
3.8
3.0
1.9
3.0
4.0
1.5
1.5
2.6
3.2
2.9
1.9
5.2
1.7
4.0
5.0
3.9
2.9
2.6
1.9
3.3
of
§
88
92
91
117
84
95
126
104
65
88
108
110
104
105
140
97
129
119
84
89
81
89
99
CIGS/DAY
18
30
25
10
25
18
0
18
40
0
0
25
20
20
0
15
50
0
0
20
0
25
28
9
1
0
0
5.3
0.4
0
1.9
0.4
0
0.6_
0
<0.1
0
1.2
0.2
0
0.5_
0
0.5
0
0
29. £
0.2
0
DUST(mpy/cf)
1
540
8
0
20
3
20
0
595
241
89
10
0
58
228
20
126
41
26
0
45
1
129
o
•H
325
20+75
1230
370
225
135+150
2390
45+70
1490
170
200
65+75
430
910
1910
785
155
225
0+70
820
51800
230
1490
2
Cn
8
2.35
1.74
1.97
1.74
1.88
2.05
1.90
2.22
1.50
1.46
1.66
1.90
2.05
1.56
1.66
1.74
1.35
2.05
2.35
2.35
2.22
1.23
1.46
O Q
row
01 CO
(J-, HI
1.2
0.1
5.2
1.8
1.0
0.5
10.5
0.2
8.3
1.0
1.0
0.3
1.7
4.8
9.6
3.8
1.0
o.e
<0.2
2.9
194.0
1.6
8.5
o — .
ro .
0) >
PM W
0.8
0.2
4.0
1.4
0.8
0.4
8.2
0.1
7.0
0.8
0.8
0.2
1.4
4.1
7.9
3.1
0.9
0.7
<0.2
2.1
140.0
1.4
7.3
DENS. (ygm/cc
0.4
0.0
1.4
0.5
0.3
0.3
3.5
0.1
5.5
0.7
0.4
0.1
0.6
2.5
1.8
2.2
0.3
0.2
0.0
1.0
75.0
0.8
2.6
MIDDLE/SIDE
0.3
-
0.3
0.4
0.3
0.3
0.3
-
0.3
0.4
0.4
-
0.2
0.4
0.2
0.3
0.3
0.4
-
-
0.5
0.4
0.2
FRONT/BACK
0.8
-
0.7
0.8
0.6
-
1.1
-
0.6
0.5
0.6
-
0.9
1.2
0.9
0.9
1.2
1.2
-
0.6
0.9
0.5
0.6
LEFT/RIGHT
0.9
-
0.9
0.9
0.9
-
1.2
1.0
1.7
0.7
0.7
0.8
0.9
1.3
1.1
1.1
0.7
0.8
1.0
0.6
1.2
0.7
0.9
o
CO
H
§
-
-
1.7
1.7
-
(2.2)
2.0
-
1.8
-
(2.1)
-
2.1
1.7
3.1
1.9
-
-
-
1.7
3.0
(1.8)
1.7
RELAXATION
(ID
-
11+
(1 + )
-
-
11+
-
11+
(III)
(11+)
-
II-
11+
11+
II-
-
11+
-
II
III+
-
II
-------
TABLE BI. DATA OF EACH SUBJECT (CON'T.)
•
0
z
E-i
U
W
5 w
5 u
in <
93 31
94 29
95 27
96 28
97 -
98 34
99 43
100 34
101 29
102 27
103 49
104 29
105 27
106 45
107 49
108 47
109 26
110 48
111 29
112 49
113 49
114 49
115 41
1
EH"
EC
U
H
3
i/y
178
174
173
175
180
171
178
172
180
169
171
175
165
166
162
172
183
173
174
175
173
167
^ *c
O> -H
* H
^^ r^
g H
u (/>
M W
g 5
75 9.5
68 9.0
107 12.0
96 11.0
87 11.5
84 10.6
69 10.2
100 11.6
60 10.0
65 8.9
78 10.2
91 10.5
92 11.2
69 10.2
85 11.0
67 10.2
61 9.6
83 10.2
83 10.0
83 11.0
74 9.5
175 14.2
68 8.9
5
.j
n
>
CJ
z
ED
^
2.9
3.7
2.8
2.2
-
3.3
4.4
2.8
3.0
4.5
2.7
2.3
2.1
2.6
1.9
2.9
3.8
3.1
3.9
2.7
3.9
-
2.6
*•"•*
k
ft ><
<#> <
* Q
ft \
< w
u e>
H
> u
125 0
111 25
104 23
95 0
28
99 255
144 10
106 4
117 0
103 10
90 0
77 (20)
91 0
20
100 0 •
90 20
119 0
97 20
117 0
97 (25)
92 25
75 0 <
87 18
•
p
iJ
10
«
><
<0.1
!._!
0
<0._1
1.0
<0.1
2.8
0.2
0
0.2_
0.8
<-o.i
0
0
<0.1
°-l
0
2.0
0
5.£
0
:°-I
0
^•^
IM
U
£
JE
f-t
u1
Cfl
3
O
2
0
0
23
-
30
19
1
22
1
267
32
2
137
378
113
2
822
1
125
596
202
7
•^*»
o^
i^
i
o
i— H
vO
W
125
760
0->80
0+150
325
0+70
1600
230
85+110
1120
1070
55+75
0+70
210
1050
1040
0+70
230
240
2500
190
0+245
150+160
V
•»_•
&
CL,
*
8
1.88
2.05
1.23
1.46
1.35
1.56
1.66
1.33
1.74
2.09
1.66
1.60
1.43
1.66
1.46
1.66
1.85
1.66
1.74
1.46
1.88
0.84
2.09
£
*r *-*
0 Q
ro W
0) W
t. H
EH fj
0.6
3.1
0
<0.9
2.0
<0.4
8.0
1.4
0.4
4.5
5.4
0.3
<0.5
1.0
6.0
5.2
<0.3
1.2
1.1
14.3
0.8
<2.8
0.6
I
h i
EH 1
0.4
2.4
<0.5
<0.8
1.8
<0.3
6.6
1.3
0.3
3.4
4.4
0.2
<0.4
0.9
5.1
4.3
<0.2
0.9
0.9
12.0
0.7
<2.4
0.5
u
CJ
^L
•
C/3
2
W
Q
0.2
0.8
0.0
0.0
-
0.0
1.8
0.5
0.1
1.0
2.0
0.1
0.0
0.4
3.2
1.8
0.0
0.4
0.3
5.3
0.2
-
0.2
w
Q
H
\
s
Q
Q
H
S
0.4
0.3
0.9
-
0.4
-
0.3
0.3
0.4
0.3
0.3
-
-
0.4
0.2
0.4
-
0.3
0.4
0.4
0.4
-
0.4
:*;
u
S
§
§
K
Cn
0.6
0.8
-
_
0.8
-
1.0
0.6
-
0.9
0.7
-
_
0.6
0.8
1.0
-
0.9
0.7
0.8
1.3
-
1.1
EH
X
u
H
«
E>
&
^
0.9
0.7
-
_
0.9
_
0.8
1.1
_
1.2
0.4
_
_
1.0
1.0
0.9
-
0.6
0.9
1.0
0.9
-
1.0
U
to
H
•>
*-*^
ft
ft
<=c
_
1.8
(1.3)
1.0
2.8
1.9
_
2.0
1.9
_
(3.9)
2.9
2.0
_
1.6
-
2.2
1.6
_
- 1
§
H
EH
3
r%
3
§
II-
IV
I
(ID
1+
III+
__
IV-
II-
_
_
(II-)
11+
11+
_
:IID
-------
The following six figures are visual displays of the distribution
data presented in Table II in the text. Our purpose is to clarify
these quantities.
o
C/)
Ql
UJ
O_
cn
UJ
m
30
20
10
0
0.2 0.4 0.6 0.8
MIDDLE/SIDE
1.0
1.2
Fig. Bl. Distribution of the ratio of the remanent field at the middle of the
torso to that on the average of both sides, for all workers who had enough Fe3O4
to yield a ratio. The sum of front and back is used for right, middle and left.
30 r
z
o
0? 20
UJ
CL
u_
o
or in
LU IU
m
^
r>
z
0
-
1 1 1 1
0.4 0.6 0.8 1.0
FRONT/BACK
1.2
1.4
1.6
1.8
Fig. B2. Distribution of the ratio of the remanent field at the front of the
torso to that at the back. The sum of the right, middle, and left is used.
The spread is quite large here because of the coarse z-spacing used, front and
back. The spread would presumably narrow if a constant z-spacing was used for
all subjects, as suggested by the correlation in Table III in the text.
42
-------
PERSONS
OJ j
o c
fe 20
o:
LJ
CD
^
i 10
n
1
1 JI _
RELAXATION GROUP
Fig. B3. Distribution of the relaxation rate for all workers who had
enough Fe304 to yield a crude relaxation curve. The groups correspond to
those of Fig. 3 in the text.
d(J
en
en
or
Q_
t in
f~j i \s
DC.
LU
CD
Z 5
O
_
-
1.0 1.4 1.8 2.2 2.6 3.0
APPARENT VISCOSITY ( L/S )
3.4
Fig. B4. Distribution of the apparent viscosity for all workers where
there was enough Fe3O4 to yield the ratio of remanent field from the long
pulse to that of the short pulse.
43
-------
NUMBER OF PERSONS
_ f\3 OJ
3 O O 0
/
0 3.2 10 32 100 320 1000
DUST EXPOSURE INDEX (LOG SCALE)
Fig. B5. Distribution of total dust exposure index. Welders are included
here. (Three workers were measured who had no indices.) A log scale is
used, except at the first interval, which goes down to zero. The workers in
this interval, mostly office workers, therefore have relatively low indices
(-0).
20 r
CO
0 15
CO
QC.
LU
CL
0 10
cr
UJ
CO
ID
^ 5
0
-
-
>0
/
"N -^
0.10 0.32 1.0 3.2 10
YEARS OF WELDING (LOG SCALE)
32
Fig. B6. Distribution of years welding for all those with any welding ex-
posure whatever. A log scale is used, except for the first interval which
extends down to but does not include zero. This interval contains many
workers who are borderline between welders and non-welders.
44
-------
Sample C. Technical Report Data Sheet, EPA Form 2220-1
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
. REPORT NO.
560/6-81-005
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
MAGNETIC LUNG
MEASUREMENTS IN RELATION TO OCCUPATIONAL EXPOSURE TO
ASBESTOS MINERS AND MILLERS OF QUEBEC
5. REPORT DATE
January, 1981
6. PERFORMING ORGANIZATION CODE
'. AUTHOFUS)
David Cohen *
Thomas S. Crowther **
(Salmi W. Gibbs * Margaret R. Bccklaka,1
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME ANO AOORESS
American Public Health Association
1015 Fifteenth Street, N.W.
Washington, D.C. 20005
10. PROGRAM ELEMENT NO.
1. CONTRACT/GRANT NO.
68-01-3859
12. SPONSORING AGENCY NAME ANO AOORESS
Office of Pesticides and Toxic Substances
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
3. TYPE OF REPORT ANO PERIOD COVERED
Final Report
4. SPONSORING AGENCY CODE
s. SUPPLEMENTARY NOTES Massachusetts Institute of Technology * ^Ctelanese
U.S. Department of Labor --- OSHA ** OteGill University 1
6. ABSTRACT
Feo04 particles (magnetic) are often attached to asbestos fibers (non-magnetic) in the primary asbestos
industries; therefore, a measurement of Fe.j04 could help determine the amount of asbestos in the lungs of
workers in these industries. As a first assessment of this method of determining retained dust, magnetic
measurements were made of the amount of Fe-j04 in the lungs of 115 miners and millers of chrysotile
asbestos. The performance of these measurements at an industrial site was found to be feasible and
practical. A relatively large amount of Fe-j04 was seen in the lungs of those with welding experience, which
masked the Fe-^O^ contributed by asbestos, therefore this group was considered separately. For the
remainder (non-welders), the amount of Fe304 was plotted against a total dust exposure index which was
available for each individual. The correlation between these quantities was not high, but was statistically
significant at the 0.01 level. For the non-smokers within that group, the correlation was higher and the
amount of Fe304 was relatively greater. These results suggest that the magnetic measurement of a
chrysotile miner and miller reflects, at least to some extent, the amount of asbestos in his lung; the scatter
could be due to individual differences in deposition and clearance, to which this measurement should be
sensitive. These results are also consistent with the possibility that less dust is deposited or retained in
smokers than in non-smokers.
7.
KEY WORDS ANO DOCUMENT ANALYSIS
DESCRIPTORS
.IDENTIFIERS/OPEN ENDED TERMS
. COSATI Field/Group
Asbestos
Occupational Exposure
Magnetic Fe304
Amount of Fe3O4 in Lung
Magnetic Dust Study
Univariate Analysis
IB. DISTRIBUTION STATEMEN1
Unlimited
19. SECURITY CLASS (This Report/'
Unclassified
21. NO. OF PAGES
2O. SECURITY CLASS (This page I
AA.
22. PRICE
r
EPA Form 2220-1 (9-73)
45
------- |