Brit. J. industr. Med., 1952, 9, 146.

CARDIOVASCULAR DISEASE IN COTTON WORKERSPART II: A CLINICAL STUDY WITH SPECIAL REFERENCE TO HYPERTENSION

BYR. S. F. SCHILLING, N. GOODMAN, and J. G. O'SULLIVAN

From the Nuffield Department of Occupational Health, Manchester University

(RECEIVED FOR PUBLICATION FEBRUARY 15, 1952)

The object of this investigation was to find outif cotton workers exposed to fine dust had a higherincidence of cardiovascular and renal disease, andin particular of hypertension, than cotton workersnot so exposed.High death rates from cardiovascular, renal, and

respiratory diseases among cotton workers havebeen recorded by the Registrar General since 1891.In a previous paper Schilling and Goodman (1951)showed that strippers and grinders, who are moreheavily exposed to fine cotton dust than any othergroup, had the highest death rate of cotton workersfor both cardiovascular-renal and respiratorydiseases.Some of the cardiovascular mortality could be

explained by methods of " book-keeping ". Before1939 the Registrar General's rules for selectingone cause of death from certificates with multiplecauses meant that many respiratory deaths wereallocated to the cardiovascular group, but when acorrection was attempted there still appeared to be areal excess mortality from cardiovascular-renaldisease among the strippers and grinders.

Clinical studies of cotton workers by Collis(1909), Prausnitz (1936), and O'Sullivan andDingwall-Fordyce (1950), and statistical investiga-tions of their morbidity by Bradford Hill (1930)and Harvey (1939) have confirmed a high incidenceof respiratory disease among strippers and grinders,but until now the only suggestion of a cardiovascularhazard is in the Registrar General's statistics.

Since 1910 methods of dust control in cottoncard-rooms have steadily improved and exposure tocotton dust has been substantially reduced. Con-ditions today are so changed for the better that therespiratory hazard may have been considerablyreduced, and the cardiovascular-renal hazard, ifever there was one, may have disappeared.'

Nevertheless, because of the interest and impor-tance ofany relationship between working conditions

and cardiovascular and renal disease, it was decidedto investigate clinically a group of cotton workers.An attempt was also made at the same time todetermine the extent of respiratory disease underpresent day environmental conditions. Specialattention was given to the incidence of " essential "hypertension, because it has been suggested thatcotton dust exposure may be a cause of its occurrenceamong cotton workers (Brit. med. J., 1947).

Method of InvestigationThe trades included in the group exposed to

dust were strippers and grinders, blow-roomworkers, carders, and under-carders. The groupnot exposed to fine cotton dust comprised weaversand warehousemen living in the same towns, andwhere possible, working in the same mills.The mills chosen were those spinning raw cotton

of the lower grades, and were in no way a represen-tative sample of the cotton spinning industry. Atthis stage of the enquiry it was more important tochoose populations likely to show positive results,so mills spinning finer grades of raw cotton andwaste cotton, where there is thought to be less dust,were not included.

There were two series of examinations series A,the first to be done, in which a group of menbetween the ages of 35 and 65 was examined, andseries B, in which a further sample between theages of 50 and 60 was examined.

Series AThe experimental and control groups were selected

as follows.Experimental Group (Card- and Blow-room

Workers).-This group comprised all men between35 and 65 who had worked in the card- and blow-rooms for at least 10 years. One hundred andforty-five men were originally selected from thefirms' books, but on more detailed examination

146

on May 24, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/B

r J Ind Med: first published as 10.1136/oem

.9.2.146 on 1 April 1952. D

ownloaded from

CARDIOVASCULAR DISEASE IN COTTON WORKERS: PART II

42 had to be rejected for a number of reasons:because they were in fact outside the age range, orhad less than 10 years' exposure to cotton dust, orhad substantial exposure to other dusts. Onehundred and three men therefore were included inthe final analysis. No man in this group refusedto be examined.

Control Group (Weavers and Warehousemen).-This group originally comprised 113 men between35 and 65, who had worked in weaving sheds orwarehouses for at least 10 years. They were selectedfrom the firms' books to form a sample matchedwith the experimental group for age and length oftime in the cotton industry. It was originallyintended to have equal experimental and controlgroups, but this was not possible. Twenty menselected from the firms' books were found ondetailed examination not to fulfil the necessarycriteria, and had to be rejected. Ninety-three menwere therefore included in the final analysis; sixweavers who refused to be examined were replacedby others from the same mills. There was noreason to believe that this refusal was related tothe problem under investigation. (Five men werein one department, and it appeared to be a groupdecision to refuse. The other man was a ChristianScientist.)

Clinical and Social Records.-All the men wereinterviewed and examined at their place of work.Occupational and family histories were recorded indetail. Clinical histories were taken with specialreference to respiratory, renal, and cardiovasculardiseases. The clinical examination was dividedinto two stages. The first stage comprised urineanalysis, the recording of certain body measure-ments, clinical examination of heart and chest, andthe recording of a " casual" blood pressure at thebeginning of the examination. The second stagewas devoted entirely to blood pressure recordings,which were made by another observer. Recordingswere taken with the patient lying down, as far as

possible in silence, after the procedure of theexamination had been explained. Systolic anddiastolic pressures were taken every three minutesand continued for 10 readings, or until the pressurestabilized at 130/70, or below. Readings weretaken to the nearest multiple of 5. The techniqueadopted in taking pressures was that recommendedby the Committee appointed by the Cardiac Societyof Great Britain and Ireland (1939). All pressureswere recorded on the left arm. Calibration at regularintervals showed no appreciable difference betweenthe two manometers used, which were of theAccosson type.

Control and experimental groups were initiallymixed and examined in random order to reducethe possibility of an external factor, such asemotional stress due to disturbance in the examina-tion room, or the effect of meals, influencing onegroup more than the other. Where both weaversand card- and blow-room workers were working inthe same mill, the observer recording pressures inthe second stage of the examination did not knowto which group the worker belonged, but it wasnot possible to make this a universal condition inthe examinations.

The Findings in Series AThe two groups were comparable in respect of

age and length of employment in the cottonindustry (see Appendix B).

Cardiovascular and Renal Disease.-The incidenceof cardiovascular and renal disease, as detected bya single clinical examination in a cotton mill, wasnot high. In the experimental group there was oneman with rheumatic heart disease and two withrenal disease. In the control group there was oneman with rheumatic heart disease and one with renaldisease (for details see Appendix A).

Blood Pressures.-The mean systolic and diastolicpressures of the experimental and control groupsare compared in Table 1. There is a statisticallysignificant difference at the 5% level between the" casual " diastolic pressures of the experimentaland control groups. The " casual " reading, how-ever, is probably the least reliable for this type of

TABLE 1

MEAN OF SYSTOLIC AND DIASTOLIC PRESSURES OF THEEXPERIMENTAL AND CONTROL GROUPS IN

SERIES A

Experimental ControlGroup Group

Casualt Lowest Casual Lowest

Number of men 100 103 92 93

Mean of systolic 141-1 127-7 141-5 127-7pressures (mm.)

S.E. ofmean .. 1.9 1-6 1.9 1-7

Mean of diastolic 87-6§ 82-0 84-7§ 80-3pressures (mm.)

S.E. ofmean . . 1.0 09 1-0 0 9

t 3 casual pressures not taken.*1 casual pressure not taken.§t=208, P=004.

147

on May 24, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/B

r J Ind Med: first published as 10.1136/oem

.9.2.146 on 1 April 1952. D

ownloaded from

BRITISH JOURNAL OF INDUSTRIAL MEDICINE

TABLE 2MEAN SYSTOLIC AND DIASTOLIC PRESSURES OF THE EXPERIMENTAL AND CONTROL GROUPS BY AGES IN SERIES A (LOWEST

PRESSURES)

Age .. .. 35- 40- 50- 60-64

Experimental Control Experimental Control Experimental Control Experimental Control

Number .. .. 12 17 50 39 31 28 10 9

Mean systolic read-ings .. 119 3 123-2 1240 125 5 135-5 130 2 132-5 138-3

S.E. of mean .. 2-9 3-6 1-8 25 3-6 3-6 5 9 5*5

Mean diastolic read-ings .. 80-0 80-3 80-8 785 84-4 81-6 835 83 9

S.E. of mean .. 15 2-3 1.1 1.1 i*8 1*6 3-7 3 1

study of blood pressures, as it may be affected bymany factors, such as apprehension about theexamination, and recent exercise. Diastolic pressuresin some people are difficult to judge unless repeatedrecordings are made. Hence this finding may havelittle or no clinical significance. All comparisonsfrom now on will be based on the lowest pressuresrecorded in the second stage of the examination.

In Table 2 the mean systolic and diastolicpressures of the two groups are compared by ages.It is only in the 50-60 age group that both themean systolic and diastolic pressures are higher forthe card- and blow-room workers than for theweavers and warehousemen, but the difference issmall and not statistically significant. Card- andblow-room workers between 60 and 64 show sur-

prisingly low systolic pressures. There is not theexpected rise with age. This finding cannot beexplained satisfactorily, but the numbers in thisage group are small, and it may be that the men

we examined were adversely selected from the pointof view of investigating hypertension. The com-

bined effects of hypertension and of the respiratorydisease, which is common among card- and blow-room workers, may have forced men with relativelymild degrees of hypertension to retire from thedusty jobs.

The age distribution of strippers and grindersand weavers at the 1931 census (more recentfigures are not available) supports this suggestion.It shows that the former group have proportionatelyfewer men over 55 than the latter (18-5% comparedwith 24%).

Hypertension.-Hemdon (1946) and Levy, Hill-man, Stroud, and White (1944), in their studies oflarge groups of apparently healthy men, regardedpressures which are consistently more than 150systolic or 90 diastolic as indicative of hypertension.We adopted a more severe standard and did notregard a man as suffering from hypertension unlessboth his lowest systolic and diastolic pressureswere at or above both the levels of 150 and 90respectively. The incidence of hypertension isshown in Table 3 by ages. The only obviousdifference between the two groups (as one wouldexpect from previous results) occurs in the 50-59age group, but the difference is not statisticallysignificant. (Clinical details of men with hyper-tension are given in Appendix A.)Thus a further group, Series B, comprising all men

in other mills between the ages of 50 and 60employed for more than 10 years, was examinedin an attempt to get a more definite result.

TABLE 3INCIDENCE OF HYPERTENSION IN AGE GROUPS IN THE EXPERIMENTAL AND CONTROL GROUPS IN SERIES A

Age .. .. 35- 40- 50- 60-64

Experimental Control Experimental Control Experimental Control IExperimental Control

Number .. .. 12 17 50 39 31 28 10 9

Lowest pressures atorabove 150/90.. 0 1 0 1 9* 3* 2 3

* *20>P>-10.

148

on May 24, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/B

r J Ind Med: first published as 10.1136/oem

.9.2.146 on 1 April 1952. D

ownloaded from

CARDIOVASCULAR DISEASE IN COTTON WORKERS: PART

Series BTwenty-eight men were examined in the experi-

mental group and 22 in the control group. A further10 men in the experimental group and 11 in thecontrol group were examined, but had to be discardedfor the same reasons as given in Series A. Nocard- or blow-room workers refused examination,but six weavers refused. Five of them, who allcame from the same mill, had decided as a groupnot to cooperate, and they were replaced by fivemen from another mill in the same area. Theother man refused because he said he was in goodhealth.

The Findings in Series BFor length of time in the industry and for age,

the experimental and control groups were com-parable (See Appendix B).

Cardiovascular and Renal Disease.-In the experi-mental and control groups there were respectivelyone man and two men with rheumatic heart disease.No man in either group was found to have renaldisease. (See Appendix A.)

Blood Pressures and Hypertension.-In Table 4the mean systolic and diastolic pressures of, andthe incidence of hypertension in, the experimentaland control groups are compared.None of the differences between the pressures of

the experimental and control groups is statisticallysignificant. The experimental group shows a higherincidence of hypertension than the control group.*

Findings for Series A and BWe will consider the two series together, as they

appear to be reasonably similar for blood pressures(see Tables 2 and 4). The data for the combinedseries are given in Table 5 and Fig. 1.There are significant differences between the

experimental and control groups for (1) the pro-portion of men with hypertension, and (2) the meansystolic pressure levels.There are similar, though not significant, differences

for the mean diastolic pressures of the combinedseries. Similar trends are found among the men

* In another mill, not planned to be included in either series,similar differences between the experimental and control groupswere also found. Men employed in this mill were examined afterthe field studies on Series B had been completed. These resultscould not be included because one stripper and grinder out of eight,and eight weavers out of 21, refused examination, and both experi-mental and control groups showed higher mean systolic and diastolicpressures than the men of this age group in other mills. At all othermills the men were seen first by one of the observers or the medicalofficer or welfare officer employed by the mill. At this mill the menhad to be seen first by the local Trade Union secretary. It was healone who explained the purpose of the investigation and persuadedthe men to cooperate. This change from the usual procedure mayexplain both the high lapse rate and the higher pressure levels, whichmay have been caused by some anxiety about the investigation onthe part of the men. The results of these examinations are given in

Appendix C.

E

TABLE 4

MEAN OF SYSTOLIC AND DIASTOLIC PRESSURES ANDINCIDENCE OF HYPERTENSION IN EXPERIMENTAL AND

CONTROL GROUPS IN SERIES B

Experi- Controlmental GroupGroup

Number of men (aged 50-59) 28 22

Mean of systolic pressures .. 140-0* 129-5*

S.E. ofmean .. .. .. -3 2-5

Mean of diastolic pressures .. 855 81-6

S.E. ofmean .. .. .. 24 2-0

Lowest pressures at or above 6 1150,90

*t = 195; P = 05 for t = 2-01.

in the rejected mill. None of these results conflictin any way, but as the levels of significance arelow it is all the more necessary to scrutinize thesefindings to see if factors not directly related tooccupation may be responsible for the differencesfound.

Observer Error.-The higher incidence of hyper-tension found among the card- and blow-roomworkers could quite simply be due to conscious orunconscious bias on the part of the investigators.There are two possible sources of observer error.First the difference could arise through one or allobservers being biased to read high for the experi-mental and/or low for the control group. It wasoriginally planned to eliminate this source of error

TABLE 5MEAN OF SYSTOLIC AND DIASTOLIC PRESSURES ANDINCIDENCE OF HYPERTENSION IN EXPERIMENTAL AND

CONTROL GROUPS IN BOTH SERIES

Experi- Controlmental GrolGroup Gru

Number of men (aged 50-59) 59 50

Mean of systolic pressures .. 137-9* 129.9*

S.E. ofmean .. .. .. 28 2-2

Mean of diastolic pressures .. 84 9t 81-6t

S.E. ofmean .. .. .. 1-4 1-3

Lowest pressures at or above 15* 4*150/90

P < *05. t P > *05.

149

on May 24, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/B

r J Ind Med: first published as 10.1136/oem

.9.2.146 on 1 April 1952. D

ownloaded from

BRITISH JOURNAL OF INDUSTRIAL MEDICINE

DISTRIBUTION OF LOWEST PRESSURESSTRIPPERS AND GRINDERS I

WEAVERS 0

PER CENT

35

30

25

20 .

15

t0

5

0 _

PER CENT

35 -

30 *

25

20

15

10

5.-

0

AGED 50 - 59 YEARS

SYSTOLIC

100 10 120 130 140 150 160 170 180 190 200 21C MM.

60 70 80 90 100 110 12

FIG. I

by ensuring that observers who tocpressures in the second stage of ti

a.hinh aawI-kth ltwuumelt r%rmecerire anz

TABLE 6THE INCIDENCE OF HYPERTENSION AND MEAN OF PRESSURELEVELS OF MEN IN 50-59 AGE GROUP WITH JOBS UNKNOWNAND KNOWN TO THE OBSERVER IN THE COMBrNED SERIES

Experimental ControlGroup Group

Job Job Job JobKnown Un- Known Un-

known known

Number of men 46 13 22 28

Number with S/D > 11 4 1 3150/90

Mean of systolic 137-9 137-7 129-5 130-2pressures

Mean of diastolic 85 6 83-8 81P6 81-6pressures

Six independent readings by observers on thesame subjects were always within a range of 5 mm.

DIASTOLIC of mercury.For the 50-59 age group observer A took 47

of the 59 pressures of the experimental group and40 out of 50 of the control group. As the proportionof pressures taken by the observer who took more

pressures than all the other observers together was

the same for both groups, it would not matterunduly if he read consistently high or low, unlesshe read high only for the experimental group or

low only for the control group. We have alreadyLr tested for the latter error in all observers and find

130 MM no evidence of it, but it is possible to test specifically

for this error in Observer A. The observed and)k the repeated expected numbers read by A, at the 150/90 level or

he examination above, for the experimental and control groups werenlllA -not lrnnw as follows:-

whether the man belonged to the experimental or

control group.This was only possible at the beginning of the

investigation, when we could find mills employingboth card- and blow-room workers and weavers atthe same place. We can, however, test for thiserror by comparing the pressures of men whosejobs were known to the observer, with the pressuresof men whose jobs were not known. Such a com-

parison is made in Table 6 for the 50-59 age group.It will be seen that there is no evidence that the-knowledge of a man's job biased the observers.Secondly, another type of observer error wouldarise if one observer read low by comparison withkother observers and took a greater proportion ofthe lowest pressures of the control group; or ifhe read high and took a greater proportion of the~experimental group's pressures.

50-59 AGE GROUP

TotalReAt or aboveatota 150/90TtlRead atbore Read byA

Number by A 150/90 Ex- Ob-Level pected served

Experimental 59 47 15 11 9 12group

Control 50 40 4 4 0 3group

There is no evidence that A was reading highfor the experimental group or low for the controlgroup.Thus observer errors are not likely causes for

the differences found between the pressures of theexperimental and control groups. The suggestion

150

on May 24, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/B

r J Ind Med: first published as 10.1136/oem

.9.2.146 on 1 April 1952. D

ownloaded from

CARDIOVASCULAR DISEASE IN COTON WORKERS: PARTH1therefore is that the difference is a real one, andwe must now examine other possible reasons for it.

Obesity and Body Measurements.-Several obser-vations have been made on the correlation betweenobesity and blood-pressure, notably those bySymonds (1923) and Treadgold (1933).

In Series A the weavers as a whole were slightlyheavier and shorter than the card- and blow-roomworkers. In series B the latter group was heavier,but the mean heights of the weavers and the card-and blow-room workers were identical. Thus it isunlikely that obesity was responsible for thisdifference.Robinson and Brucer (1940) showed that body-

build, as indicated by the ratio of chest circumferenceto height, was closely correlated with hypertension,and that obesity occurred most frequently in thelateral or broad-built types. They concluded thatbody-build, regardless of weight, determines in agreat measure the predisposition of any individualto hypertension and that the role of obesity inhypertension is small. Among the men in the50-59 age group, the grades of body-build describedby Robinson and Brucer were similarly distributedamong the experimental and control groups.

There was no evidence that the higher incidenceof hypertension among the card- and blow-roomworkers was due to anthropometric differencesbetween the groups.

Heredity.-Studies of family histories of patientswith hypertension by Platt (1947), Weitz (1923),Hines (1937), and Ayman (1934) have shown thatheredity is a factor of importance in the aetiologyof " essential " hypertension. Our groups mustnow be examined to see if this result can be explainedby heredity.The method described by Platt (1947) for classi-

fying family histories into four groups, positive,probable, incomplete, and negative, was adopted.In this study 42% of all the men gave incompletefamily histories, whereas in Platt's series of 187cases this proportion was only 21%. A possibleexplanation for this difference is that Platt's patientswere ill enough to attend hospital and were probablyaware of their complaint, whereas the men in ourseries were fit enough to work and no one of themwas aware that he had hypertension. It has beenstressed by Doll and Buch (1950) and others, thatthe patient who knows he has a particular diseaseis more likely to be interested in searching outsimilar cases in his family than the patient who isnot suffering from, or does not know he has, adisease.With such a high proportion of incomplete

family histories we cannot say to what extent

heredity influenced our result, although there wasno evidence that it had done so. Card- and blow-room workers did not appear to have a strongerfamilial tendency to hypertension than weavers.Each group had 22% of men with positive familyhistories of hypertension.

In this series the family history did not appearto influence the blood pressure significantly. Menwith positive and probably positive family historieswere grouped together and compared with the menwith negative family histories. An analysis ofvariance showed that for the systolic pressuresonly the variation between occupational groupswas significant, but it was not so between historiesfor either the systolic or diastolic pressures, nor wasthere any evidence of a significant interactionbetween histories and occupational groups.

Renal Disease.-There was no difference in theincidence of renal disease found in the two groups,but the clinical methods used had to be crude,and it is possible that the higher incidence ofhypertension in the card- and blow-room workersmight be caused by renal disease which we failedto diagnose.

If we compare the medical histories of the twogroups (all ages) for diseases such as nephritis,trench fever, and scarlet fever, the experimentaland control groups respectively had 18 and 23 menwith suggestive histories of renal disease. It there-fore seems unlikely that hypertension secondary toundetected renal disease was the cause of thedifference between the two groups.

Industrial Respiratory Disease.-The incidenceand severity of industrial respiratory disease(byssinosis) among the card- and blow-room workersexamined must be considered, as it may bear somerelation to the incidence of hypertension.

Sixty-three per cent. of the men were found tobe complaining of tightness of their chests althoughonly 11% showed any serious disability which wasmost likely to be due to emphysema.

Pneumoconiosis and emphysema are importantaetiological factors in pulmonary hypertension(Rosenthal, 1930; McMichael, 1948; Gelfand,1951), but their aetiological importance in systemichypertension or the effect of a raised pulmonarypressure on the systemic pressure has not beenclearly established. Gelfand found in cases of corpulmonale that the systemic pressure was usuallynormal. Parkinson and Hoyle (1937), however,found that systemic hypertension was one of thediseases most commonly associated with emphysemain a series of patients attending hospital-; --but thesepatients may have sought- tr u r theiremphysema because of the added- coiplication of

151

on May 24, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/B

r J Ind Med: first published as 10.1136/oem

.9.2.146 on 1 April 1952. D

ownloaded from

BRITISH JOURNAL OF INDUSTRIAL MEDICINE

INCIDENCE AND GRADE OF BYSSINOSIS AMONG THE CARD-AND BLOW-ROOM WORKERS*

Number Per-Grade of Men centage

0 No symptoms .. .. 45 34

I Complaining of tightness 52 40of the chest on Mondaysonly; symptoms notprogressive

II Above symptoms getting 16 12progressively worse andoccurring on days otherthan Mondays. Dis-appearance of symp-toms when not exposedto dust

III Complaining of chest 15 11tightness on every work-ing day and consideredto have some permanentdisability with effortintolerance when not atwork

Not graded because of 3 2incomplete histories

Total .. 131

* Byssinosis does not -occur among weavers and warehousemen inGreat Britain. Here all stages of the industrial respiratory diseaseof card- and blow-room workers have been included under the termbyssinosis. Byssinosis, because of its definition for compensation,has often been used only for the disabling stage of the disease.

systemic hypertension which was not necessarilyrelated to emphysema. Heimann, Castberg,Walters, David, and Meyn (1950) in a study offoundrymen in Illinois found that significantlyelevated systemic blood pressure levels were relatedto increasing degrees of pulmonary fibrosis. Inthe present investigation, of the men between50 and 60 in the experimental group, there were17 men who had no evidence of industrial respiratorydisease. Their mean pressures were 131 8 mm.systolic and 81-8 mm. diastolic and three of themhad hypertension. The mean pressures of the41 men with industrial respiratory disease were140 0 systolic and 86 1 diastolic,* and 12 of themhad hypertension. The slight differences in themean pressures and in the incidence of hypertensionare not statistically significant; the numbers aresmall and the reliability of the methods of diagnosingand grading byssinosis is by no means certain.Byssinosis in its earlier stages can at present only

be diagnosed by the occupational and medicalhistories (Schilling, 1950), and it is probable that inthis study there was some error in both the diagnosisand the grade. By adopting more objective methodsof assessing pulmonary disability in card- andblow-room workers, it is hoped in a later investiga-tion to get a more reliable assessment of therelationship between byssinosis and systemic hyper-tension, and also to test the reliability of the clinicalgrading of byssinosis.

DiscussionThe difference found in the incidence of hyper-

tension among card- and blow-room workers andweavers is not big enough to explain the largedifference in their mortality rates shown by theRegistrar General for 1921-23 and 1930-32. Thereare explanations for this discrepancy.

It has already been shown that some of themortality excess was due to the allocation ofrespiratory deaths to the cardiovascular group ofdiseases. The Registrar General's mortality figuresrelate to periods 20 and 30 years ago, when therewas much more dust in the card-rooms than thereis today. Although we found a substantial incidenceof industrial respiratory disease in the older men,and it may well have been a hang-over from thepast, it was less than the incidence 40 years agowhen Collis (1909) found 90% of men of all agesaffected.As the cotton mills in which this investigation

was made had no records of the men who hadleft, we were only able to study the population fitfor work. Our data cannot, thererore, give anaccurate picture of the true incidence of hypertensionin the occupational groups studied. This, however,will not invalidate the relative incidence of hyper-tension in the two trades unless hypertension ismore likely to cause unfitness and disability forwork in one trade than in the other throughdifferences in physical demands and environmentalconditions. It is known that many people withbenign hypertension can live a normal working lifefor many years without symptoms, if they are keptin ignorance of their condition and not preventedfrom living a normal life by their medical advisers(Naish, 1946). The physical demands of work incard- and blow-rooms and weaving sheds are moreor less the same, but the combined effects oncard- and blow-room workers of respiratory diseaseand hypertension might well make the latter diseasemore disabling for them than for weavers. If thisdoes happen, it would decrease our chance of findinghypertensives among the card- and blow-roomworkers we examined. In this respect it may besignificant that we found so few card- and blow-

* Men with Grade III byssinosis had mean pressures of 144-4/87-5mm.Men with Grades I and II byssinosis had mean pressures of

138-9/85 8 mm.

152

on May 24, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/B

r J Ind Med: first published as 10.1136/oem

.9.2.146 on 1 April 1952. D

ownloaded from

CARDIOVASCULAR DISEASE IN C07TON WORKERS: PART II

room workers with hypertension in the 60-65 agegroup.

It is not possible to state categorically thatheredity does not account for the higher incidenceof hypertension among the card- and blow-roomworkers, because of the incompleteness of thefamily histories. There is, however, a suggestionthat hypertension may be related to industrialpulmonary disease. This or some other occupationalinfluence may be the cause of our findings.

If this study had been made 30 years ago whenthe card-rooms of British cotton mills were verydusty and if the data on cardiovascular mortalitycould have been analysed more accurately, theresults might well have been more conclusive. Itwould be of interest to know if there is similarbut more definite evidence of an occupationalhypertension among cotton workers in the othercountries where dust suppression methods in cottonmills have been even less successful. Nevertheless,in this country, in spite of the efforts to reducedust, this investigation has shown that industrialrespiratory disease still remains a serious problemamong the older men employed in mills spinningthe coarser grades of cotton.

SummaryThis paper describes a clinical study in Lancashire

cotton mills spinning and weaving the coarsergrades of cotton of card- and blow-room workers,who are exposed to fine cotton dust, and weavers

and warehousemen who are not so exposed. Specialattention was given to the incidence of hypertension.Repeated blood pressures were taken during theexamination. Men were considered to be sufferingfrom hypertension when their lowest pressureswere at or above both levels of 150 mm. of mercuryfor the systolic pressure and 90 mm. of mercury forthe diastolic pressure.

In the age group 50-59, 15 out of 59 (25%h)card- and blow-room workers had hypertension as

defined, whereas four (8%) out of 50 weavers andwarehousemen had hypertension. This differenceis statistically significant at the 5%' level.There was no evidence that this was due to

observer error, to anthropometric differences betweenthe two groups, or was related to a higher incidenceof renal disease among the card- and blow-roomworkers. Nor was there definite evidence to suggestthat the difference was due to a greater familial

tendency to essential hypertension among card-and blow-room workers than among other cottonworkers.

Sixty-three per cent. of the card- and blow-roomworkers had symptoms of byssinosis and there was

a suggestive relation between industrial respiratorydisease and hypertension, which will be investigatedmore fully.

We should like to thank the many employers andworkers whose cooperation and interest made this studypossible, and in particular Mr. Henniker Heaton, ofthe Federation of Master Cotton Spinners' AssociationsLtd., Mr. A. Roberts, of the Card and Blowing RoomOperatives and Ring Spinners' Association, Mr. A.Naismith of the Amalgamated Weavers' Association;Dr. J. N. Morris of the Social Medicine Research Unitof the Medical Research Council for much advice andcriticism ; Professor Lane and Professor Robert Plattand other colleagues inside and outside the departmentfor their help and advice. We would also like to thankDr. J. Gregory and Dr. R. F. L. Logan for con-

siderable help with the examinations; Dr. Roscoe forexamining the samples of urine; and Dr. R. Murray andMr. W. Moore of the Factory Department of theMinistry of Labour and National Service for helpingus to select the cotton mills.

REFERENCESAyman, D. (1934). Arch. intern. Med., 53, 792.British Medical Journal (1947). 1, 727.Cardiac Society of Gt. Britain and Ireland (1939). Brit. Heart J.,

3, 261.Collis, E. L. (1909). Annual Report of the Chief Inspector of

Factories and Workshops for 1908. Home Office [Cd. 4664].H.M.S.O. London.

Doll, R., and Buch, J. (1950). Ann. Eugen., Lond., 15, 135.Gelfand, M. L. (1951). Amer. J. Med., 10, 27.Harvey, P. N. (1939). Appendix to the Report of the Departmental

Committee on Compensation for Card Room Workers, p. 19.H.M.S.O., London.

Heimann, H., Castberg, H. T., Walters, F. J., David, W. D., andMeyn, A. W. (1950). U.S.P.H.S. Public Health ServicePubl. No. 31, p. 55 seq.

Herndon, R. F. (1946). An Introduction to Essential Hypretension.Springfield, Illinois.

Hill, A. Bradford (1930). Ind. Health Res. Board Rep., No. 59.H.M.S.O., London.

Hines, E. A. (1937). Ann. intern. Med., 11, 593.Levy, R. L., Hillman, C. C., Stroud, W. D., and White, P. D. (1944).

J. Amer. med. Ass., 126, 829.McMichael, J. (1948). Edinb. med. J., 55, 65.Naish, J. M. (1946). Bristol med.-chir. J., 63, 106.O'Sullivan, J. G., and Dingwall-Fordyce, I. (1950). The Incidence of

Byssinosis in the Waste Cotton Industry. Personal com-munication.

Parkinson, J., and Hoyle, C. (1937). Quart. J. Med., n.s. 6, 59.Platt, R. (1947). Ibid., n.s. 16, 1 1 1.Prausnitz, C. (1936). Spec. Rep. Ser. med. Res. Coun., Lond.,

No. 212.Robinson, S. C., and Brucer, M. (1940). Amer. J. med. Sci., 199, 819.Rosenthal, S. R. (1930). Arch. Path., Chicago, 10, 717.Schilling, R. S. F. (1950). Brit. med. Bull., 7, 52.

, and Goodman, N. (1951). British Journal ofIndustrial Medicine8, 77.

Symonds, B. (1923). J. Amer. med. Ass., 80, 232.Treadgold, H. A. (1933). Lancet, 1, 733.Weitz, W. (1923). Z. klin. Med., 96, 151.

153

on May 24, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/B

r J Ind Med: first published as 10.1136/oem

.9.2.146 on 1 April 1952. D

ownloaded from

>7

0 GoC..m1a .

C4

o."

>0 0

r,-

04~

4o4P-.4Y

~-I0n003

W

I sz

>0

.00

.4

10>0r.

r_.

I^

I

I

1-.0

c0

.4

c

0'e

>0

CL

0-

4)

.e

14I 0

Gn

.0

>.cl

I.'Ax0

.4

.(

00

10o.0I

ipwo

ICZ41

>0.0

.0

co

>l

0la

>0

r.0.0

014

141

I.2

a

la

..

0

141

0

0

r.

.200

a

Z

1-4

laD

0

(A

0

.2I

.0

la0

1-

0>0-

b.0

1.

0

.4

0

0

>010X

es

.2.-olce2S.

IEx

II.0

>06I ^

.201c

i<

1 5 1IDj>4D|I E E EE E E E E E E E

~~~~~~ 0

0 ~~~~4~~~4.)~~~~ 0 0 0 0 0 0 0 0 0 0o-l~Z_-t 1<U Z Z ZIZZ z z ZZz z.1

~~~~~~~~~~~~~~~~~~~~~~~~~W0 (O000 N

0 000WV 0 a-0

00 tn~~~~~~~0~W0 W

21o1° v} o WI£ o Y:v) Iv) £ EV,IC) oW) WI00 r~- -e0m°0=7I>

O_ _ _______

CZ ' CAe - - .a c Y

0e ¢ I 1s to0 tE0~~~~~~~

4) 4a 4)

>1 I~~~~~~lI ~

__ lzO Z Iz zol> 0._ oD 0 0D

oo - Fe IT-|6}oi-oFxo-> __>}1

08bee1 Iwa 'AYusr r r wY 1e 1Y 1I Is III 4)1Q IQI 14) 114)vI

z z z P6 0. C6 44 44 0. 0- O.*-

to 'n_ II 00

0 s- 11

X 1

0< I~0IW-I0".I 0

Im ImI0o ON.en e

ImI,It m0

.42(A

.0

.5'U)

0la

0

r_0

I.4rAI 0(A

>0>0.04

.0

C64

0.~

4)'4)14)2

4)1

zCu

4)

I-4)

U i.0

... :

4). __

0)F-0z0)

U

m

PC

inll%

oO.

z CII

0o0o

-ILIl o

on May 24, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/B

r J Ind Med: first published as 10.1136/oem

.9.2.146 on 1 April 1952. D

ownloaded from

0

0

._*0

0

Cu;N

0D

0C11

._

.f0,.

=

0.

.00.

0

CZ

.2

0

Iw;

0

e.00

IoU I_

.0

U)Cd

._

I..

Cu*0

._

2

CO)

._Cd

0 0 0 0

z z alZ° Iz

.0.0 .~~~~~~0.0Cu Cu 00 CZo o IU

o o 0X o

O o 0)o~~~oo I

.

Cu

._

z

I..oo

00

oCl

0C

0.cO

co Cu

Z I ZtOa to

121 °

*0

0-to Ca0

ON

CuI.20

E

00

CIIto

.0

0I 0O Cu

00

z

o

'0

la

0

C9

1-1m

au

IsS

0

'U

0

0toV

a;t

t3

r.e0

:0

0

.°

0

0.4

ICuI00t._

0

1'MCCu

aslo100

0.:0%

0

:o

._

c

st

I>10.

C)o0

S.

Cut.

e:s

IC.-

I x

._

>0

e

.0._0Cu

06

0

._

0

C

cc500._*0.

0._C Cow

a

0

0.6.

>0w

0

0

0.

>0

Cd

*0.0

.01

r-I a:

I :1:

*0Cd

la

._

Es

a..2

.0cce,C)

0

+ 0.

2 22c,< E - 2E- 2E - 0 2'E 'E'EEE

D D D D D DS0Cu D D E0 .0

0 0 0 0 0 0 0C) o0 0 0

z z z zoz z z z z zV

FWoF% 0 0 0u V) £ F ^

°-1°- j°F-°X °ll 00o 0 0 0 00 0

Y E 0X SY 0 Cu y

to a0: 3 E E y E - °Cu0 a3~ ~~~01

0 .uCuo C.t u ~ ~ t c).

~~~~ 0

z 0

e.a

0

1°

-~:Cu

00

z

a

0

E

I c1-

0

0

Cu

0

0

aL

0

Eoc

1-4

0.

E0

0

440

c

E000

S.C's

0.

E0

0

0

0C)C

m00

00- :t.

Cu

Cu

to0

z

4)

00

z

I .t

I 1.

ko

%U)C)

00zoz

O0

0.0

to

Cd0U0

.0

0

o0to0In

*0

0.

._

¢-

0 C-4 0 00 C4 t- 0 N 0tn tn un W) "o en 1%0 .o 't

1.

1 3:

°< --oas.< <I_

on May 24, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/B

r J Ind Med: first published as 10.1136/oem

.9.2.146 on 1 April 1952. D

ownloaded from

BRITISH JOURNAL OF INDUSTRIAL MEDICINE

APPENDIX B

COMPARISON BETWEEN EXPERIMENTAL AND CONTROLGROUPS FOR AGE AND LENGTH OF TIME IN THE COTTON

INDUSTRY

1.-Series A

Experimental ControlAge Group Group

402}(400o) 17f (37%)

45- 21V 22 (41%)50- 13f (33 16f (41°055- 18 1 (270%) 12 (2301 )60-64 lOf 9f (230

Totals 103 93

Years inIndustry

10 6 (6%o) 11 (12%0)20 36 (35%//) 31 (33%)30 43 (42%O) 38 (41%/ )40 1 8 (17%' ) 1 1 (12%/)50 - 2 (2%o)

Totals .. 103 93

2.-Series B

Time in Cotton IndustryAge ____ (Years)

<20 20- 30- 40- Total

Experimenital group50-54 - 2 12 - 1455-59 1 7 6 14

Total - 3 19 6 28

Colntrol gr'Ooup50-54 .. 1 9 - 1055-59 2 - 3 7 12

Total .. 2 1 12 7 22

APPENDIX CINCIDENCE OF HYPERTENSION AND MEANS OF LOWESTPRESSURES OF EXPERIMENTAL AND CONTROL GROUPS

IN MILL NOT INCLUDED IN SERIES A AND B

Experi- Cotlmental Control

Number of men ......... . 7 13No. above 150/90 level 3 3Mean systolic pressure .. 152-9 135 4S.E. of meal? .. .. 17.4 4.7Mean diastolic pressure .. 100 0 82-7S.E. of mean .. .. 117 30

156

on May 24, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/B

r J Ind Med: first published as 10.1136/oem

.9.2.146 on 1 April 1952. D

ownloaded from