British Journal of Industrial Medicine 1982;39 368-374

Dust exposure and mortality in an American factory
using chrysotile, amosite, and crocidolite in mainly
textile manufacture
A D McDONALD,' J S FRY,' A J WOOLLEY,2 AND J C McDONALD2

From the Department ofEpidemiology, ' St Mary's Hospital Medical School, and the TUC Centenary Institute
ofOccupational Health,2 London School ofHygiene and Tropical Medicine, London, UK

ABSTRACr This report describes the second in a series of three parallel cohort studies of asbestos
factories in South Carolina, Pennsylvania, and Connecticut to assess the effects of mineral fibre
type and industrial process on mortality from malignant mesothelioma, respiratory cancer, and asbes-
tosis. In the present plant (in Pennsylvania) mainly chrysotile, with some amosite and a small
amount of crocidolite, were used primarily in textile manufacture. Of a cohort of 4137 men
comprising all those employed 1938-59 for at least a month, 97% were traced. By the end of
1974, 1400 (35%) had died, 74 from asbestosis and 70 from lung cancer. Mesothelioma was
mentioned on the certificate in 14 deaths mostly coded to other causes. All these deaths occurred
after 1959, and there were indications that additional cases of mesothelioma may have gone
unrecognised, especially before that date. The exposure for each man was estimated in terms of
duration and dust concentration in millions of dust particles per cubic foot (mpcf) from available
measurements. Analyses were made both by life table and case referent methods. The standar-
dised mortality ratio for respiratory cancer for the whole cohort was 105*0, but the risk rose
linearly from 66.9 for men with less than 10 mpcf.y to 416.1 for those with 80 mpcf.y or more.
Lines fitted to relative risks derived from SMRs in this and the textile plant studied in South
Carolina were almost identical in slope. This was confirmed by case referent analysis. These
findings support the conclusion from the South Carolina study that the risk of lung cancer in
textile processing is very much greater than in chrysotile production and probably than-in the
friction products industry. The much greater risk of mesothelioma from exposure to processes in
which even quite small quantities of amphiboles were used was also confirmed.

Cohort mortality studies were conducted simultane-
ously in three American plants to clarify the effect of
mineral type of asbestos and of processing, as dis-
tinct from production, on mortality from malignant
mesothelioma, lung cancer, and asbestosis. Prelimi-
nary reports describing results by duration of emp-
loyment in all three plants' and detailed exposure-
response findings in one of them-a textile plant
that used only chrysotile2-have been published.
This report concerns a second plant which manufac-
tured mainly textiles but also friction products and
packings, many of which were made from the textile
products. In addition to chrysotile, some amphibole
asbestos (amosite and crocidolite) was also used. In

Received 15 February 1983
Accepted 7 March 1983.

the third plant friction materials and packings were
made from chrysotile only and no textiles were
manufactured.3 The plant dealt with in the present
report has been the subject of two previous
studies45; in both, excess mortality was reported
from chronic respiratory disease, lung cancer, and
malignant mesothelioma. In neither study was
exposure estimated other than by duration of
employment. In the present study the cohort was
much larger, and individual dust exposures were
estimated from environmental measurements made
in the work place.

Materials and methods
The workforce at the plant averaged 1200 men and
women between the late 1930s and 1975; before
then it was rather smaller. Social security numbers

368

 o
n
 A

p
ril 5

, 2
0

2
1

 b
y g

u
e
st. P

ro
te

cte
d

 b
y co

p
yrig

h
t.

h
ttp

://o
e
m

.b
m

j.co
m

/
B

r J In
d

 M
e

d
: first p

u
b

lish
e

d
 a

s 1
0

.1
1

3
6

/o
e
m

.4
0
.4

.3
6
8
 o

n
 1

 N
o
ve

m
b
e
r 1

9
8
3
. D

o
w

n
lo

a
d
e
d
 fro

m
 

http://oem.bmj.com/


Dust exposure and mortality in an American factory using chrysotile, amosite, and crocidolite
were allocated in 1937. Details of all those em-
ployed before 1 January 1959 for at least one calen-
dar month, with a recorded social security number,
were submitted to the United States Social Security
Administration. Those whose numbers and names
were consistent with the administration's records
comprised the cohort. The cohort therefore fell into
two groups: (a) all those first employed after 1937
and (b) those first employed before that date.
Because the latter group was selected in that it
excluded employees who had left before 1938, the
two groups were analysed separately. In fact, the
findings were similar for those two groups and are
reported together in this paper.

In all, 4137 men and 998 women met the cohort
criteria. We ascertained survival status as of 31
December 1977 through local inquiries and from
information on deaths provided by Social Security
Administration. All but 113 men (3%) and 61
women (6%) were traced; of those traced, 1400
men (35%) and 170 women (18%) had died. Copies
of death certificates were obtained from state
departments of vital statistics or from the country
where death had occurred for 1354 (97%) of the
men and 165 (97%) of the women. In two male
deaths without certificate the cause was ascertained
from other sources. Causes of death were coded
according to the seventh revision of the Interna-
tional Classification of Disease (ICD) by a single
qualified nosologist. This paper deals only with the
1392 male deaths, shown by age and cause in table
1. Eight deaths are omitted from this table because
of insufficient information on age. Female mortality
in this and the other two plants will be reported
separately.

THE PROCESS
The plant was opened in a rural area of central
Pennsylvania in the early 1900s. From the beginning
it has manufactured a variety of products, almost all
containing asbestos. The factory had a large asbestos
textile division of traditional type. Crude fibre was
cleaned, opened, carded on modified wool carding
machines, spun using mule and later ring-type
frames, and wound and woven into cloth or plyed to
form tape. Much of the cloth or tape was treated
with resin, dried, and then finished into brake lin-
ings, clutch facings, and transmission strips. Since
the early days the plant has also manufactured fric-
tion materials by blending raw asbestos with resin-
binding ingredients followed by extrusion, rolling, or
bricketting.

Chrysotile, obtained mostly from Canada and
Rhodesia, was the main type of asbestos processed,
with some 3000 to 6000 tons of varying quality used
annually. Crocidolite and amosite were used from
1924 onwards for making insulation blankets for
locomotives and turbines and equipment for chemi-
cal factories, and paper mills. Amosite insulation
blankets were produced in large quantities during
1942-5 to meet US naval specifications. Crocidolite
packings were also produced, as were crocidolite
and amosite filter materials. Much of the crocidolite
was imported as yarn and only about three to five
tons of raw fibre were used annually whereas, in
1943, the use of amosite reached a peak of 600 tons.
From 1925 to 1931 manufacture of insulation mat-
tresses and filter cloth was transferred to a disused
opera house in the town. The mattresses were
packed with crocidolite or amosite, the crocidolite
yarn being imported from England. After 1931, the
opera house was used as a warehouse and the manu-

Table 1 Male deaths by age and certified cause

Cause of death (ICD code) Age at death Total

<45 45-64 ¢65

All causes 191 667 534 1392
Malignant neoplasms:
Lung* (162-164) 3 49 18 70
Oesophagus and stomach (150-151) 1 7 6 14
Colon and rectum (152-154) 2 21 12 35
Other abdominal* (155-159) 3 16 5 24
Larynx (161) 0 0 0 0
Other* (140-148, 160, 165-205) 16 57 40 113

Heart disease (400-443) 43 285 245 573
Respiratory tuberculosis (001-008) 5 4 2 11
Other respiratory (470-522, 525-527) 6 17 25 48
Pneumoconiosis (523-524) 2 48 24 74
Cerebrovascular (330-334) 3 33 44 80
Accidents (800-999) 74 44 20 138
Other known causes* 23 73 80 176
Cause not known 10 13 13 36

*In 13 cases in these categories, mesothelioma was given as the cause of death; in one death ascribed to asbestosis, mesothelioma was also
mentioned.

369

 o
n
 A

p
ril 5

, 2
0

2
1

 b
y g

u
e
st. P

ro
te

cte
d

 b
y co

p
yrig

h
t.

h
ttp

://o
e
m

.b
m

j.co
m

/
B

r J In
d

 M
e

d
: first p

u
b

lish
e

d
 a

s 1
0

.1
1

3
6

/o
e
m

.4
0
.4

.3
6
8
 o

n
 1

 N
o
ve

m
b
e
r 1

9
8
3
. D

o
w

n
lo

a
d
e
d
 fro

m
 

http://oem.bmj.com/


McDonald, Fry, Woolley, McDonald

Table 2 Estimated average prevailing dust concentrations (MPCF) in main departments 1930-70

1930
Textile:

Preparation 15-0 * 14-6
Carding 12-0 * 9-3
Spinning 7.0
Twisting 7*2
Winding 3-0
aoth weaving 7-9
Tape weaving 7.3
Felted tape 2-0
Rope 4 0

Friction:
Woven brakes
Extruded brakes
Dry mix brakes
autch
Brake finishing
Sanding and finishing
Finishing and shipping

Packings, gaskets
Maintenance, etc

1940

* 3-5
3-2

* 2-5
* 4-0
* 1-5

*3-4 *
*3-1 *

* 1-0
* 2-5

2-0 1-5
2-0 1-5

10-0 6-0
2-0 1-5
2-0 1-5
2-0 1-5
0-5
1.3 1-3
0-5 0-5

1950

* 18

1-4
0o9

1960

2-0 *
*2-0 * 1-6
* 1.8 *

*0-5 * 1-1
0-8

* 1-2
0-5
1-2 *

1-0
1-0
4-0
1-5
1-0
1-0
0-2
0-6
0-2

0-7
0-8
3-8
1-5
0-7
0-7
0*2
0-7
0-2

*Asterisks shown against textile processes indicate approximate date of improvements usually associated with technical change. Figures for
friction and other departments are estimates for each decade.

Table 3 Age at start, duration ofemployment, and dust exposure (male only)

Length ofgross service

<1 1, <5 5, <20 ¢'20 Total

No 1248 906 855 1013 4022*
Average age at start (years) 28.80 29-30 30-77 27-22 28-92
Gross servce (years) 0-40 2-39 11-01 30-63 10-71
Net service (years) 0.38 1-87 8-06 27-51 9-18
Average dust concentration (mpcf) 2-60 2-40 2-73 1-58 2-32

*Excluding two whose employment histories were incomplete.

facture of cloth and mattresses was returned to the
main site.

EXPOSURE
Conditions are reported to have been very dusty in
the 1920 and 1930s, although oil spraying intro-
duced to facilitate handling probably had some
damping effect. The first steps to reduce exposure
were taken in 1930; these included wetting,
improved handling methods, and ventilation hoods.
By about 1939 exhaust ventilation had been instal-
led in the textile mill with substantial reduction in
dust levels; thereafter improvement was gradual.

Reports of surveys made by the Metropolitan Life
Insurance Company in the period 1930-9 were
available, as were surveys by the United States
Public Health Service made in 1967 and in 1970.
Measurements were made routinely by the company
from 1956 onwards. Until 1967, all measurements
were made by impinger and recorded in millions of
particles per cubic foot (mpcf). Dust level estimates
for each department were made by the hygienist of
our group (AJW) and are summarised in table 2.
These estimates do not take into account certain

additional exposures which may have been quite
short but which were certainly very heavy. These
were associated with daily "blowing down" and
"whipping the burlap" in the dust house. At
weekends waste asbestos thus recovered in the dust
house was swept up and bagged for recycling. A new
dust collection system introduced in the early 1950s
virtually removed this hazard.
The data available on exposure are summarised in

table 3, grouped by duration of employment. The
average age at start of work and the average dust
concentration differed little in the four duration
groups.

STATISTICAL METHODS
Mortality in the cohort was analysed (a) by the
man-years life table method of Hill,6 using Pennsyl-
vania death rates for reference, and (b) by the log
rank procedure of Mantel and Heanszel,' exactly as
in the South Carolina study.2

Results

The standardised mortality ratio (SMR) for all

1970

1-5

15

0-6

370

 o
n
 A

p
ril 5

, 2
0

2
1

 b
y g

u
e
st. P

ro
te

cte
d

 b
y co

p
yrig

h
t.

h
ttp

://o
e
m

.b
m

j.co
m

/
B

r J In
d

 M
e

d
: first p

u
b

lish
e

d
 a

s 1
0

.1
1

3
6

/o
e
m

.4
0
.4

.3
6
8
 o

n
 1

 N
o
ve

m
b
e
r 1

9
8
3
. D

o
w

n
lo

a
d
e
d
 fro

m
 

http://oem.bmj.com/


Dust exposure and mortality in an American factory using chrysotile, amosite, and crocidolite 371
Table 4 Male deaths 20 years after first employment, by cause, in relation to length ofservice

Cause of death * Length ofgross service years

<1 1, <5 5, <20 a20 Complete cohort

0 SMR 0 SMR 0 SMR 0 SMR 0 SMR

All causes 171 87-2 154 106-2 187 104-5 383 127-2 895 109-0
Malignant neoplasms:

Respiratory 9 69.6 3 32.9 14 128-8 27 158-9 53 105-0
Abdominal 8 72-9 11 133.7 11 105-9 24 131-3 54 112-7
Other 19 132-4 16 152-0 15 118-5 32 155-3 82 141-1

Heart disease 77 92-7 77 125-1 78 100-2 153 115-7 385 108-5
Respiratory tuberculosis 0 - 1 133-4 0 - 2 67-3 3 51-7
Other respiratory: 4 54-2 2 38-1 11 161-0 50 442-4 67 215-0
Pneumoconiosis (2) - (1) - (10) - (46) - (59) -

Cerebrovascular 7 54-9 10 106-5 10 77-7 20 87-6 47 81-2
Accidents 13 117-5 15 181-1 8 87-2 9 60-1 45 103-5
Other known 30 75-2 15 52-4 37 103-0 62 103-1 144 87-2
Not known 4 4 3 4 15

*As in table 1 except that ICD codes 160-164 are here grouped under "respiratory" malignnt neoplasms and the "other respiratory"
category includes only bronchitis, pneumonia, and pneumoconiosis (ICD 490-502, 523-4).

Table 5 Male deaths 20 years after frst employment, by cause, in reladon to dust exposure (mpcf.y) accumulated to 10
years before death

Cause of death * Dust exposure (mpcf.y)
(See table 4)

<10 10 <20 20 <40 40 <80 380

0 SMR 0 SMR 0 SMR 0 SMR 0 SMR

All causes 470 93.1 86 82-1 130 1256 105 174-9 104 215-2
Malignant neoplasms:

Respiratory 21 66-9 5 83.6 10 156-0 6 160-0 11 416-1
Abdominal 26 90 2 8 130-5 5 79-7 8 218-8 7 237-2
Other 47 130-4 5 68-5 11 148-6 7 164.7 12 372-8

Heart disease 221 102-7 41 89-2 60 130*6 34 130-5 29 108-5
Respiratory tuberculosis 1 34-3 0 - - - 1 169-7 1 163-6
Other respiratory: 8 43-6 5 122.0 10 263.0 14 623-3 30 1689-2
Pneumoconiosis (4) - (1) - (9) - (9) - (36) -

Cerebrovascular 27 78-3 1 13-3 10 133-5 8 187-2 1 29-3
Accidents 33 12041 3 56-2 1 18-6 6 193-9 2 91-0
Other known 74 73.3 17 80-0 23 109-7 21 172-2 9 97-8
Not known 12 1 0 0 2

Table 6 Dust exposure in male deaths from selected causes and controls (Mantel-Haenszel analysis8)

Dust exposure (mpcfy) accumulated up Chi square
to IO years before death ofcase Difference Linearity

<10 10<20 20<40 40<80 ;80

Pneumoconiosis (ICD 523):
Deaths 3 4 10 11 28 39-56 39-17
Expected 14-6 8-1 10.9 8-1 14-3
Relative risk 1 4.04 13-72 14-93 37-90

Lung cancer (ICD 162-4)
Deaths 20 4 10 6 11 5-77 4-98
Expected 24-4 5-2 8 0 5-6 7-7
Relative risk 1 0-83 1-54 2-90 6-82

Abdominal cancer (ICD 150-9):
Deaths 26 8 5 8 7 3-22 1-09
Expected 28-8 6-8 7-0 5-3 6-1
Relative risk 1 1-15 0.66 2-45 2-85

All causes
Deaths 451 81 121 100 99 34-66 26-12
Expected 476-6 104-5 118-6 80-4 72-0
Relative risk 1 0-82 1-20 1-6 2-12

 o
n
 A

p
ril 5

, 2
0

2
1

 b
y g

u
e
st. P

ro
te

cte
d

 b
y co

p
yrig

h
t.

h
ttp

://o
e
m

.b
m

j.co
m

/
B

r J In
d

 M
e

d
: first p

u
b

lish
e

d
 a

s 1
0

.1
1

3
6

/o
e
m

.4
0
.4

.3
6
8
 o

n
 1

 N
o
ve

m
b
e
r 1

9
8
3
. D

o
w

n
lo

a
d
e
d
 fro

m
 

http://oem.bmj.com/


372

causes of death was 109-0. Those employed for
under one year had an SMR of 87*2, and those who
had worked 20 or more years, 127.2 (table 4).
Malignant neoplasms, heart disease, and "other
respiratory" disease were mainly responsible for the
higher SMR in these long term workers. The other
respiratory category included bronchitis and
pneumonia (ICD 470-502) and pneumonoconiosis
(ICD 523-4) and was chosen for study because
expected figures for pneumoconiosis alone were not
available. Table 5 shows SMRs by cause and by
accumulated dust exposure. The SMR for all causes
rose steadily from 93 1 for men with an exposure at
under 10 mpcf.y to 215*2 in the highest category (-
80 mpcf.y). Respiratory, abdominal, and other
malignant disease and the non-malignant other
respiratory group all contributed to this rising trend.
On 14 death certificates a diagnosis of mesothelioma
was specified: 10 were pleural tumours and four
peritoneal. These deaths occurred in the period
1960-75. One (in 1960) was 16 years after first
employment; the remaining 13 occurred 25-53
years after first employment. Two of the deaths from
mesothelioma had been given the ICD code 199
(malignant neoplasms of other and unspecified
sites); another 30 deaths 15 more years after first
employment were given the code 199. Seventeen of
these 30 deaths occurred before 1965, the year after
which most of the deaths from mesothelioma occur-
red. The diagnosis given in many of these cases was
consistent with an unrecognised peritoneal
mesothelioma.
The Mantel-Haenszel (log rank) analysis (table 6)

bore out the exposure-response relationships
observed in table 5. There is a small shortfall (5%
overall) between the numbers of cases used in this
analysis and in the man-years analyses presented in
tables 4 and 5. The deficiency is explained by failure
to find matching controls for every selected case.

McDonald, Fry, Woolley, McDonald

Discussion

The present cohort in the Pennsylvania plant was
constituted in exactly the same way as that in the
South Carolina chrysotile textile plant described
elsewhere.2 Compared with the South Carolina em-
ployees (see table 3), those in Pennsylvania had
worked for similar periods but were on average
some four years older when they started work
(28.93 years compared with 25*08 years). The
Pennsylvania cohort was exposed to a somewhat
higher average dust concentration: 2-32 mpcf com-
pared with 1.80 mpcf in South Carolina. The mortal-
ity pattern in Pennsylvania resembled that in South
Carolina in showing a rising SMR with increasing
dust exposure for all causes of death, for respiratory
cancer, and for pneumoconiosis. For respiratory
cancer, however, the SMR for the lowest exposure
group (less than 10 mpcf.y) was 115-5 in South
Carolina but only 69-9 in Pennsylvania. By contrast
with South Carolina, where the SMRs tended to be
above 100 for causes unrelated to asbestos, and for
all causes in very short term employees,2 the oppo-
site was true in Pennsylvania. It seems likely that in
both cohorts lack of comparability with the relevant
state populations may be the explanation. Having
regard for this possibility, the use of relative risks is
perhaps more appropriate than SMRs for comparing
the respiratory cancer mortality of the two cohorts.
Table 7 shows that the relative risks of death from
all causes, respiratory cancer, and pneumoconiosis
in the two plants were extraordinarily similar. In
both cohorts the relationships of respiratory cancer
to exposure were essentially linear (figure) with
slopes that were nearly identical (South Carolina,
RR = 1 + 0-059 mpcf.y; Pennsylvania, RR = 1 +
0*051 mpcf.y).
A conceivable explanation for depression of the

SMRs in the Pennsylvania plant was that the em-

Table 7 Relative risks based on SMRs by cumulative exposure in two plants

mpcf.y
<10 10 <20 20 <40 40 <80 380

All causes:
South Carolina plant 1-0 1-09 1-36 1-48 2-29
Pennsylvania plant 1.0 0-88 1-35 1-88 2-31

Respiratory cancer
South Carolina plant 1.0 1-28 2-13 2-93 7-21

(1-32) (1-68) (2-80) (3-86) (949)
Pensylvania plant 1-0 1-25 2-33 2-39 6-22

(1-26) (1-58) (2-94) (3.03) (7.87)
Bronchitis, pneumonia, and

pneumoconiosis:
outh Carolina plant 1.0 1-81 6-40 21-36 19-67

Pennsylvania plant 1.0 2-79 6-03 14-29 38-74

Figures in italics are relative risks calculated from SMRs at zero exposure derived from fitted line.

 o
n
 A

p
ril 5

, 2
0

2
1

 b
y g

u
e
st. P

ro
te

cte
d

 b
y co

p
yrig

h
t.

h
ttp

://o
e
m

.b
m

j.co
m

/
B

r J In
d

 M
e

d
: first p

u
b

lish
e

d
 a

s 1
0

.1
1

3
6

/o
e
m

.4
0
.4

.3
6
8
 o

n
 1

 N
o
ve

m
b
e
r 1

9
8
3
. D

o
w

n
lo

a
d
e
d
 fro

m
 

http://oem.bmj.com/


Dust exposure and mortality in an American factory using chrysotile, amosite, and crocidolite

RR -

* South Carolina /
g _ 0 Pennsylvania

8 0 "

6

5 -

4 -

/,"
3 -- 0

2 -

1
50 100 150

npcf y

Relative risk ofrespiratory cancer and accumulated dust
exposure in two mainly textile plants. (Lines fitted by FI)K
Liddell using the methods ofHanley and Liddell.)

ployees had smoked fewer cigarettes than the general
population of the state,,perhaps because of the high
proportion of Mennonites in the local community.
We ascertained, however, that in fact very few Men-
nonites were employed in the plant. Smoking his-
tories had been recorded in a standard manner for
all those at work in 1978 or later in the two plants
and also in the plant in Connecticut where the third
cohort was studied.3 Of men born in 19109 (who
would have been included in the three cohorts) the
proportions who had never smoked were: 25% (of
36) in Pennsylvania, 11% (of 95) in South Carolina,
and 16% (of 206) in Connecticut. Thus the propor-
tion of non-smokers (25%), though somewhat
higher than in the other two cohorts, is based on
very small numbers and no firm conclusion can be
drawn.

Similar proportions of all deaths in the two
cohorts were from malignant disease (17% in South
Carolina and 18% in Pennsylvania) but the types of
malignancy differed. In South Carolina respiratory
cancer accounted for 47%, abdominal 25%, and
other types 28% whereas in the Pennsylvania plant
the corresponding proportions were reversed, 27%,
29%, and 44%. Moreover, in South Carolina no
systemic relationship with exposure was seen for
abdominal or other types of malignant disease
whereas in Pennsylvania there was evidence of such
a relationship.
The increased risk of mesothelioma in the Penn-

sylvania plant (14 cases in 1392 male deaths (1%)
compared with one case in 867 (0.1%) in South
Carolina) raises the question of whether the abdom-
inal and more particularly other types of cancer
included undiagnosed cases of mesothelioma. There
is some support for this idea in the substantial
number coded to ICD 199 (malignant disease of
other and unspecified sites) and the fact that 17 of
these deaths occurred before 1964 when malignant
mesothelioma started to become more generally
recognised. Once again we find evidence in this
study of the special risk of mesothelioma associated
with exposure to even quite small proportions of
amphibole, in this case predominantly amosite.
The very similar exposure-response relationships

for respiratory cancer and asbestosis observed in this
and the South Carolina plant support our previous
conclusion that the risks of these diseases in
chrysotile production (mining and milling) and in
textile manufacture are quite different. In the third
plant studied, a friction materials plant in Connec-
ticut, there was little or no excess risk of respiratory
cancer or asbestosis.3 This was also true in a friction
materials plant in the United Kingdom.9 Possible
reasons for the striking epidemiological
differences-fibre size distributions in particular-
have been discussed elsewhere.210

Many people and organisations helped in most
aspects of this research. We thank especially Sheryl
Nixon for data processing and Roslyn MacNish for
coding causes of death. A grant toward the cost of
the study was made to St Mary's Hospital Medical
School by the Institute of Occupational and
Environmental Health of the Quebec Asbestos Min-
ing Association. Facilities for completing the study
were given to Dr A D McDonald by the Institut de
Recherche en sante et en sdcuritd du Quebec.

Requests for reprints to: Dr J C McDonald, School
of Occupational Health McGill University, 1130
Pine Avenue West, Montreal, Canada H3A 1A3.

References

' McDonald AD, Fry JS. Mesothelioma and fibre type in three
American asbestos factories. Preliminary report. Scand J Work
Environ Health 1982;8 supple:53-8.

2 McDonald AD, Fry JS, Woolley AJ, McDonald JC. Dust expos-
ure and mortality in an American chrysotile textile plant. Br J
Ind Med 1983;40:361-67.

3McDonald AD, Fry JS, Woolley AJ, McDonald JC. Dust
exposure and mortality in an American chrysotile friction
products plant. Br J Ind Med (in press).

Mancuso TF, Et Attar AA. Mortality pattern in a cohort of
asbestos workers. JOM 1967;9: 147-62.

5Robinson C, Lemen R, Wagoner J. Mortality patterns

373

 o
n
 A

p
ril 5

, 2
0

2
1

 b
y g

u
e
st. P

ro
te

cte
d

 b
y co

p
yrig

h
t.

h
ttp

://o
e
m

.b
m

j.co
m

/
B

r J In
d

 M
e

d
: first p

u
b

lish
e

d
 a

s 1
0

.1
1

3
6

/o
e
m

.4
0
.4

.3
6
8
 o

n
 1

 N
o
ve

m
b
e
r 1

9
8
3
. D

o
w

n
lo

a
d
e
d
 fro

m
 

http://oem.bmj.com/


374

1940-1975 among workers employed in an asbestos textile
friction and packing products manufacturing facility. In:
Lemen R, Dement JM, eds. Dusts and diseases. Park Forest
South, Illinois: Pathotox Publishers, Inc, 1979:131-43.

6 Hili ID. Computing man years at risk. Br J Prev Soc Med
1972;26: 132-4.

Mantel N, Haenszel W. Statistical aspects of the analysis of data
from retrospective studies of disease. J Natl Cancer Inst
1959;22:719-48.

McDonald, Fry, Woolley, McDonald
* Peto R, Pike MC. Conservation of the approximation E (0-

E)2/E in the log rank test for survival data or tumor incidence
data. Biometrics 1973;29:579-84.

'Newhouse ML, Berry G, Skidmore JW. A mortality study of
workers manufacturing friction materials with chrysotile asbes-
tos. Ann Occup Hyg 1982;26:899-909.

'0 McDonald JC. Aspects of the asbestos standard. In: Gee JBL,
Morgan KC, Brooks SM, eds. Occupational lung disease. New
York: Raven Press, 1983.

 o
n
 A

p
ril 5

, 2
0

2
1

 b
y g

u
e
st. P

ro
te

cte
d

 b
y co

p
yrig

h
t.

h
ttp

://o
e
m

.b
m

j.co
m

/
B

r J In
d

 M
e

d
: first p

u
b

lish
e

d
 a

s 1
0

.1
1

3
6

/o
e
m

.4
0
.4

.3
6
8
 o

n
 1

 N
o
ve

m
b
e
r 1

9
8
3
. D

o
w

n
lo

a
d
e
d
 fro

m
 

http://oem.bmj.com/