Occupational and Environmental Medicine 1994;51:557-563

Emphysema and airway obstruction innon-smoking South African gold miners withlong exposure to silica dust

Eva Hnizdo, Gerhard K Sluis-Cremer, Eugene Baskind, Jill Murray

AbstractObjective-Occupational exposure tosilica dust is associated with significantimpairment oflung function. The presentstudy investigates which pathologicalchanges in the lung are associated withimpairment oflung function in silica dustexposed workers who were life-long non-smokers.Methods-242 South African white goldminers who were lifelong non-smokersand who had a necropsy at death werestudied. The pathological features identi-fied at necropsy were the degree and typeof emphysema, the presence of airwaydisease, and the degree of silicosis in thelung parenchyma and pleura. These fea-tures were related to lung function testsdone a few years before death, to type ofimpairment (obstructive or restrictive),and to cumulative silica dust exposure.Results-The degree of emphysemafound at necropsy was not associatedwith a statistically significant impair-ment of lung function or with dust expo-sure. The degree of silicosis in the lungparenchyma and the large airways dis-ease (based on mucus gland hyperplasia)were associated with a statistically signif-icant impairment of lung function. Thelarge airway disease was, however, notpositively associated with dust exposureor silicosis. In miners with a moderate ora higher degree of limitation of airflowthe main findings were silicosis, heartdisease, and obesity. The presence ofsmall airways disease could not be estab-lished from the necropsy material.Conclusion-The results indicate that thelevel of exposure to silica dust to whichthese miners were exposed, without aconfounding effect of tobacco smoking, isnot associated with a degree of emphy-sema that would cause a statistically sig-nificant impairment of lung function.Silicosis of the lung parenchyma wasassociated with loss of lung function.Other factors that may play a part inimpairment of lung function in theseminers are obesity and heart disease.

(Occup Environ Med 1994;51:557-563)

It has been shown in experimental animals'and also in humans2-4 that exposure to silicaor mineral dust containing a large percentageof silica can result in pathological changes in

the lung parenchyma other than silicosis andthat these changes result in impairment oflung function. The changes are emphysema,5 6thickening of small airways,' and pronouncedfibrosis and pigmentation of the respiratorybronchioles.2

It is uncertain whether the impairment oflung function in workers exposed to silica dustis due to the change in the alveoli (emphy-sema) or in the small airways. The AmericanThoracic Society suggests that in subjects atrisk of developing chronic obstructive pul-monary disease (COPD), pathologicalchanges in the peripheral airways precede thedevelopment of emphysema and that thesechanges on their own, without emphysema,may be responsible for subtle abnormalities inpulmonary function tests that are not associ-ated with physical impairment.7

Workers in South African gold mines areexposed to low levels of respirable dust ( % 0 4mg/m3) containing a high percentage of crys-talline silica (- 30%). Epidemiological studieshave shown that exposure to silica dust is pos-itively associated with the degree of emphy-sema, assessed in paper mounted lungsections obtained at necropsy,56 but that theestimated effect of silica dust relative to thatof smoking is of smaller magnitude.6 It hasalso been shown that the centriacinar type ofemphysema occurred mainly in men whosmoked, whereas in the non-smokers therewas a small degree of the panacinar type ofemphysema.6 Other studies have found a sta-tistically significant association between expo-sure to silica dust and loss of lung function(FEVy, FVC, FEFM,75%) in smoking andnon-smoking gold miners.48 In white goldminers, the impact of tobacco smoking andthe combined effect of tobacco smoking andsilica dust were found to be the most impor-tant risk factors contributing to disability andmortality from COPD.'The purpose of the present study was to

investigate further which pathological changesfound in workers exposed to silica dust aremainly responsible for the loss of lung func-tion. As the impact of tobacco smoking onCOPD has been shown to be substantiallygreater than that of silica dust, only gold min-ers who were lifelong non-smokers wereincluded in this study. In particular, answersto two questions were sought. Firstly, in non-smoking gold miners with airflow limitation,which of the following can be implicated: sili-cosis, emphysema, airway disease, or otherfactors? Secondly, does the assessment of air-way disease, as made at necropsy on South

Epidemiology Branch,National Institute ofEnvironmental HealthSciences, MDI A3-05,PO Box 12233,Research TrianglePark, North Carolina27709, USAE HnizdoEpidemiologyResearch Unit,Medical Bureau forOccupationalDiseases, PO Box4584, Johannesburg2000, South AfricaG K Sluis-CremerE BaskindNational Centre forOccupational Health,PO Box 32492,Braamfontein 2017,South AfricaJ MurrayCorrespondence to:Dr Eva HnizdoAccepted for publication8 April 1994

557 on S

eptember 2, 2021 by guest. P

rotected by copyright.http://oem

.bmj.com

/O

ccup Environ M

ed: first published as 10.1136/oem.51.8.557 on 1 A

ugust 1994. Dow

nloaded from

Hnizdo, Sluis-Cremer, Baskind, Murray

African gold miners, (bronchitis, based onmucus gland hyperplasia, and bronchiolitis,based on goblet cells metaplasia) reflect lungfunction impairment associated with exposureto silica dust? The morphological changesexamined were (a) emphysema (type anddegree); (b) small airways disease (goblet cellsmetaplasia); (c) large airways disease (mucuscell hyperplasia); and (d) the presence anddegree of silicosis in lung parenchyma andhilar glands.

Black gold miners, who have much lowertobacco consumption, could not be used inthis study as they have a substantially lowernecropsy rate and less reliable exposurerecords than the white miners.

Materials and methodsSELECTION OF STUDY SUBJECTSBy South African law it is obligatory for anattending physician to remove the cardiorespi-ratory organs of a deceased subject who isknown or suspected of having worked inmines and to send them for an examination tothe Medical Bureau for OccupationalDiseases (MBOD), provided that the next ofkin agrees. If the subject dies within a 100 kmradius of Johannesburg, the whole body issent for a full or partial necropsy. The patho-logical examinations are performed at theNational Centre for Occupational Diseases(NCOH) according to a standard method setdown by the Pneumoconiosis Act of 1956.The results of the pathological examination ofthe lungs are used for compensation purposes.Since 1974, all the necropsy records havebeen stored on a computerised database(PATHAUT).The study subjects were selected from

PATHAUT if they fulfilled the following cri-teria: (a) full or partial necropsy examinationduring the years from 1974 to 1990; (b) whiteminers with predominantly gold mining expo-sure (those with asbestos exposure or morethan two years of mining on mines other thangold mines were excluded); (c) age at death45 years or more; (d) a lifelong non-smoker.Of the 16 724 necropsies performed during1974-90 on white miners, 5681 were done onminers who had predominantly gold miningexposure and whose lungs were inflated atnecropsy. Of these, 712 were coded onPATHAUT as non-smokers and 510 as ofunknown smoking habits. The MBOD med-ical files of all these miners were examined tocheck their smoking history; smoking habitsare recorded annually since 1960. At the endonly 242 miners met the criteria and remained

Table 1 Agreement between study scores and originalNCOH scores for the degree ofemphysema

Study NCOH pathologistsscores(%) None Insignificant Moderate Marked Total

< 10 125 31 1 0 15710-34 28 50 3 0 8135-64 0 2 2 0 4>65 0 0 0 0 0Total 153 83 6 0 242

in the study. Most men excluded because ofmissing smoking data stopped their miningexposure before 1960. In other excluded min-ers, the medical file indicated past smoking.About 86% of deceased white gold miners

have a necropsy at NCOH. To establishwhether non-smokers have the same necropsyrate as other miners, we have examined thenecropsy rates for non-smokers included in a17 year mortality follow up study done on2260 gold miners.8 Of the 239 non-smokers,69 died and 87% of these had a necropsy atNCOH, but not all had inflated lungs. Thenecropsy rates were 77%, 94%, and 63% inthose with no compensation, first degree, andsecond degree compensation in life, respec-tively. In gold miners compensable diseasesare pneumoconiosis, chronic obstructive pul-monary disease (COPD), tuberculosis, andprogressive systemic sclerosis. The seconddegree is usually awarded to miners who havepneumoconiosis and COPD). Thus thedegree of compensation in life affects to someextent whether a miner has a necropsy.

ASSESSMENT OF EMPHYSEMAAt full or partial necropsy the lungs areinflated with formaldehyde and papermounted whole lung sections are made fromone lung, usually the right one, with theGough-Wentworth technique. The degree ofemphysema is assessed for compensation pur-poses by assigning a score between 0 and100%. For our study, the type and degree ofemphysema were reassessed by an experi-enced NCOH pathologist who was blind tothe men's smoking or dust exposure historyand the purpose of the study. The gridmethod with 20 radiating zones was used.Each zone was assigned scores for centriacinar(CAE) and panacinar (PAE) type of emphy-sema in such a way that the sum of the twoscores in each zone ranged from 0 to 5. Thesum score over all 20 zones was calculated forCAE, PAE, and total emphysema (TE =CAE + PAE), and each was expressed as apercentage of 100 (or less than 100 if somezones were missing).The emphysema scores in our study were

categorised and compared with the originalNCOH grades given as none (<10%),insignificant (10-34%), moderate (35-64%),and marked (> 65%) (table 1). The disagree-ment between none and insignificant degree isnot as important as the agreement in thatthere were only two cases with a moderatedegree of emphysema (with scores of 37%and 45%). The disagreements on insignificantand moderate degree were for the cases withemphysema scores close to 35%.

ASSESSMENT OF AIRWAY DISEASE AT NECROPSYThe pathological assessment of airways dis-ease is done on lungs that are already at vary-ing stages of decomposition when preservedin formaldehyde. This compromises theassessment of the large airways disease, buteven more so of the small airways disease, asonly small loose pieces of bronchiolar epithe-lium can be seen. Nevertheless, a moderate

558 on S

eptember 2, 2021 by guest. P

rotected by copyright.http://oem

.bmj.com

/O

ccup Environ M

ed: first published as 10.1136/oem.51.8.557 on 1 A

ugust 1994. Dow

nloaded from

Emphysema and airway obstruction in non-smoking South African gold miners with long exposure to silica dust

degree of bronchitis or bronchiolitis, based onnecropsy diagnosis of large and small airwaysdisease, in addition to a given number of yearsof service in gold mines, is a sufficient crite-rion for compensation for COPD in SouthAfrican gold miners.

Small airways disease is assessed mainly bygoblet cell metaplasia in bronchioles (airwaysunder 2 mm in diameter). It is graded,according to the ratio between the number ofgoblet cells and the total number of cells seen,as insignificant (up to 50%), moderate(50-75%), and marked (>75%). Large air-ways disease is assessed by mucus glandhyperplasia in a main bronchus. It is gradedaccording to the ratio between mucus glandsand total bronchial epithelium thickness, withthe Reid Index, as insignificant (<6), moder-ate (6-7), and marked (>7).The repeatability of the diagnosis of goblet

cell metaplasia and mucus gland hyperplasia,and of the assessment of the quality of thematerial, was tested on a sample of 24 sub-jects (of the 242) who had varying degrees ofobstructive impairment of lung function. Oneexperienced pathologist made the assessmentindependently twice. For mucus gland hyper-plasia, the repeatability of the diagnosis wasstatistically significant (Spearman correlationcoefficient = 070, p = 00004). Even so, ofthe seven classified as moderate on the firstreading (none was classified as marked), threewere classified as insignificant on the secondreading. The agreement on quality of thematerial for the assessment of mucus glandhyperplasia was relatively good, as 19 out ofthe 24 were classified as good on both read-ings; the rest were poor or unreadable. Forgoblet cell metaplasia, the repeatability waspoor and the quality of the material was classi-fied as good only in eight subjects on the firstreading, and half of these were classified aspoor on the second reading. The rest wereclassified as poor or unreadable. Because ofthe unreliability of the diagnosis of goblet cellmetaplasia, only the presence and degree ofmucus gland hyperplasia was used in a statisti-cal analysis (two subjects were excluded as onthe original NCOH reading the material wasclassified as poor).

SIIUCOSIS AT NECROPSYThe presence of silicosis is established forcompensation purposes by a macroscopic andmicroscopic examination. The degree of sili-cosis in the lung parenchyma is based on theprofusion of palpable islets in the lungs as fol-

Table 2 Distribution ofstudy subjects according to number oflungfunction tests done atMBOD

No Age atfirst Age at last Years ofNo of of test test Age at death mining CDItests subjects Mean (range) Mean (range) Mean (range) Mean Mean

0 68 63 (46-92) 18 20-51 57 58 (33-76) 66 (45-84) 21 27-42 34 57 (38-67) 62 (49-70) 67 (49-85) 24 34-83 23 56 (45-71) 65 (57-75) 68 (60-77) 26 40-14 13 55 (45-72) 64 (56-75) 68 (62-78) 26 38-05 15 56 (45-67) 67 (59-75) 70 (59-86) 30 45-26 32 53 (41-72) 66 (50-81) 68 (51-82) 29 47-3Total 242

lows: none (0), insignificant (<5), slight (5-9),moderate (10-29), and marked (,>30).Where there is massive fibrosis (confluentislets measuring >2 0 cm in diameter) this iscategorised as marked. The microscopicexamination of histological slides is used toconfirm that the nodules are silicotic. Thedegree of fibrosis in the lymph nodes is gradedas some are fibrosed, are fibrosed and denselyfibrosed. The degree of silicosis of the lungparenchyma and pleura and the degree offibrosis in the lymph nodes are recorded in thePATHAUT file. The assessment of the pres-ence of silicosis is considered generally to bereliable. In the 242 miners the degree of sili-cosis was as follows: 123 had none, 61 had aninsignificant, 32 had a slight, 17 had a moder-ate, and nine had a marked degree of silicosisof the lung. There were 68 and 63 minerswith a moderate and marked degree of lymphnodule fibrosis respectively.

ASSESSMENT OF CHRONIC FUNCTIONALAIRFLOW IMPAIRMENTOf the 242 miners, 174 miners had lung func-tion tests done at MBOD. Only miners whorequest disability compensation, or who seekexamination for health reasons, have lungfunction tests done. Table 2 shows the distrib-ution of the 242 miners according to the num-ber of lung function tests they had done atMBOD, their average age at the first and mostrecent test, years of gold mining, and cumula-tive dust exposure. The miners who did nothave any tests were on average younger atdeath and had significantly lower dust expo-sure in comparison with the others. Severalfactors are involved in the selection processfor having lung function tests, but a whiteminer with clinical signs of impairment ismore likely to have tests than a miner withshorter exposure who feels healthy.The highest forced expiratory volume in

one second (FEV,), forced vital capacity(FVG), and slow vital capacity (VG) (selectedfrom spirogram or flow-volume curve), as wellas height and weight, were coded from eachlung function test recorded in the miner'sMBOD medical file. Based on results of lungfunction tests miners are categorised, at thetime of testing, as normal, mild obstructive,moderate obstructive, marked obstructive,and restrictive depending on whether theirlung function measurements fall within thelower 90% confidence interval (90% CI) ofthe predicted value, within the 90% and 95%CI, within the 95% and 99% CI, and abovethe 99% CI respectively. If a miner is cate-gorised in the obstructive category, furthertests are made to ascertain whether the airflowreduction is reversible after inhalation ofbronchodilator or in time, to exclude asthmaor other reasons for the abnormal tests. Thefinal diagnosis is recorded with the tests andwas used in our study to identify miners withimpairment of lung function.

SILICA DUST EXPOSUREExposure to mining dust was coded for eachminer from the official service records in

559 on S

eptember 2, 2021 by guest. P

rotected by copyright.http://oem

.bmj.com

/O

ccup Environ M

ed: first published as 10.1136/oem.51.8.557 on 1 A

ugust 1994. Dow

nloaded from

Hnizdo, Sluis-Cremer, Baskind, Murray

Table 3 Characteristics of the 242 miners

Characteristics Mean (SD) (range)

Year of birth 1915 (9 4) (1890-1940)Year of death 1981 (3 8) (1975-1989)Age at death 65-9 (8 6) (45-92)Years of mining 23-1 (11 0) (1-48)First year in dust in mines 1940 (9 9) (1909-1974)Last year in dust in mines 1970 (9 5) (1939-1988)Cumulative dust (mg/m'-years) 6 8 (2 4) (0-5-20 2)

terms of the number of shifts worked in eachoccupation in a mine. A cumulative dustexposure in mg/m3-years was calculated as asum of products of the number of shifts spentin an occupational category and intensity ofdust exposure in that category as estimated byBeadle9 (for details see Hnizdo and Sluis-Cremer'°). This measure of dust is stronglyrelated to risk of silicosis.'0

STATISTICAL ANALYSISAs the emphysema scores were not normallydistributed, non-parametric as well as para-metric tests were applied for comparisons. Toidentify factors associated with lung functiontests, degree of emphysema, and degree ofmucus gland hyperplasia, the multiple linearregression model and the stepwise method ofvariable selection were applied.

ResultsTable 3 shows the characteristics of the min-ers. On average, the miners were 65-9 years ofage at death and had 23 (range 1-48) years ofnet gold mining service during 1940-70.

EMPHYSEMA AND RELATED FACTORSMainly, an insignificant degree of panacinaremphysema type was found. The figure showsthe distribution of emphysema scores for thetotal, centriacinar, and panacinar emphysematypes. The median emphysema score for totalemphysema was 5 (range, 0-45%). Only foursubjects were classified as having a moderate

CentriacinarM Panaciner

Total emphysema

0 4 8 12 16 20 24 28 32Emphysema score (%)

36 40 44 48

degree of emphysema (score 35-65%). Ofthese, two had normal lung function tests(FEV,% > 70), one miner had an FEV,% of51, but had coronary heart disease, and oneminer, who had only one year of undergroundservice, was 87 years old at death and had pul-monary tuberculosis. When included in themultiple regression as an independent vari-able, the degree of emphysema was notselected as a predictor of the most recent lungfunction measurements (p < 0-15) whenother confounding factors were adjusted for(table 4). Neither was the degree of emphy-sema associated with years of gold mining,cumulative dust index, or parenchymal silicosiswhen age at death was adjusted for. Thedegree of panacinar emphysema, however,was found to be associated with the degree ofhilar gland nodules (p < 0 05). When age atdeath and year of birth were adjusted for, themean emphysema scores (SE) increased withincreasing degree of hilar gland fibrosis as 5-2(1.03) for none, 7-2 (1-30) for insignificant,8-2 (1-01) for moderate, and 9 05 (1-09) formarked degree. The differences between themean emphysema scores for moderate andmarked categories and the lowest categorywere statistically significant (p < 0-05).

BRONCHITIS DIAGNOSED AT NECROPSYOf the 240 subjects in whom the diagnosis ofmucus gland hyperplasia was done, 27 (11%)had a moderate and none had a markeddegree. In the multiple regression analysis,when the degree of mucus gland hyperplasiawas entered as a response variable, theexplanatory variables positively associatedwere age at death (p = 0 03) and body massindex (BMI; weight(kg)/height(m)2; p =0 08). The degree of silicosis of the lungparenchyma was negatively associated (p =0 005). Silica dust was not selected into theregression model.

LUNG FUNCTION TESTSTable 4 shows the factors associated with themost recent lung function measurements (atp < 0 15). The factors not selected into theregression model were silicosis of the pleuraand hilar glands nodules, years of gold min-ing, and cumulative dust exposure. Thedegree of bronchitis was the second mostimportant predictor of FEVy and FVC afterage at test, then came height, silicosis of thelung, and BMI. The BMI and height were notrelated to FEV,%. None of the dust exposurevariables were selected into the model, evenwhen silicosis was not considered. This maybe due to the selection process, by which theminers with low dust exposure were excludedfrom lung function testing, resulting inreduced variation in the dust exposure vari-able. The degree of emphysema was notrelated to any of the tests.

There were 73 miners whose FEV1/FVCratio was at some time less than 0 70. Medicalfiles of these miners were examined indepen-dently by a specialist physician (GKS-C) and aphysician experienced in interpretation of

(5

crCT

LL

Frequency distribution for the centriacinar, panacinar, and total emyphysema type scoresfor non-smokers.

560

on Septem

ber 2, 2021 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.51.8.557 on 1 August 1994. D

ownloaded from

Emphysema and airway obsmion in non-smoking South African gold miners with long exposure to silica dust

Table 4 Factors associated with lungfincion test values (n = 174): regressioncoefficients (SE)

Lwmgfunction tes

Factor FEV,(l) FVC(l) FEVIFVC%

Age at test -0-022 (0-007)*** -0-021 (0-007)*** -0-166 (0-1 13)Mucus gland hyperplasia -0-279 (0-077)** -0-223 (0-079)*** -3-110 (1-281)**Height (cm) 0-030 (0-008)** 0-048 (0-008)***Lung silicosis -0-102 (0 041)* -0-092 (0-047) -1-833 (0-683)*Body mass index -0-025 (0-01 1)* -0-031 (0-01 1)*

*p<0-05; **p < 0-01;* p < 0-001.

pulmonary function (EB) to select minerswith an appreciable degree of airflow limita-tion. According to a physician's report on thelast medical examination, 43% of the minerswere obese, 62% had high blood pressure,and 56% had heart disease in life. In 17 of the73 miners, the comment on the lung functiontest indicated an obstructive or restrictivepattern. The medical records of these men

were extracted and table 5 summarisesselected medical diagnoses made in life, andat necropsy. It is noteworthy that the necropsyfindings for COPD,-namely, emphysemascore, mucus gland hyperplasia, and gobletcell metaplasia, do not indicate substantialpathological changes consistent with COPDin most of these men. Also in life, only oneminer (No 12) had clinical evidence ofchronic obstructive lung disease. There were

six miners who had a history of chronic bron-chitis in life; however, the necropsy findingsfor mucus gland hyperplasia and goblet cellmetaplasia do not reflect this condition (thismay also be due to unreliability of these twonecropsy findings). Of the six men with a

diagnosed restrictive pattern, five were shownto have some degree of silicosis at necropsy.Obesity and heart disease were diagnosed in10 of these men. In comparison with the rest ofthe miners, the 17 miners had significantlyhigher cumulative dust exposure up to 35years of age (3-5 v 2-2 mg/m3-years, p <0-0007), from 35 to 45 years of age (3-1 v 1-9mg/m3-years, p < 0-0002), and total cumula-tive dust up to death (10-4 v 6-6 mg/m3-years,p < 0-0001).

DiscussionIn this study of lifelong non-smokers exposedto silica dust, only an insignificant degree ofpredominantly panacinar emphysema was

found. Silica dust exposure or parenchymalsilicosis were not related to the degree ofemphysema. The degree of panacinar emphy-sema, however, was related to the degree offibrosis in hilar glands, which could haveacted as a surrogate for exposure to silicadust. The degree of emphysema was notfound to be a significant predictor of the mostrecent lung function tests (table 4). Thus theresults indicate that the amount of emphy-sema found in lifelong non-smokers exposedfor many years to silica dust is unlikely tocause a moderate degree of airflow limitation.In smokers, however, the association betweensilica dust and the degree of emphysema was

statistically significant, suggesting thattobacco smoking potentiates the effect of sil-ica dust.6 A considerable amount of lungdestruction has to occur, however, beforemacroscopic emphysema becomes identifi-able.The necropsy material used in the present

study was not suitable for a diagnosis of thick-ening and fibrosis of small airways which, ithas been suggested, is caused by silica dustexposure and associated with a significant lossof lung function.'2 Nevertheless, it was theintention of this study to identify miners clas-sified by lung function tests as having moderateairflow limitation. Because the degree ofemphysema was known, it was hoped that onecould deduce indirectly whether airway dis-ease or emphysema was responsible for theobstructive pattern. On examination of themedical files of the 73 miners whoseFEVI/FVC ratios were less than 0-70, it wasfound that obesity and general lack of fitnesscould also have contributed to the airflow lim-itation." Table 5 summarises the clinical andnecropsy findings for the 17 miners whosemedical files indicated the presence of some

impairment of lung function. Of the six menwith a restrictive pattern, five had somedegree of silicosis at necropsy. Emphysema

Table 5 Comparison of medicalfindings in lfe and at necropsy in 17 South African gold miners diagnosed as having respiratory impairment in life

Findings at necropsyFindings in life

SilicosisCase Functional x Ray film EmphysemaNo FEVI (%) pattern Bonchitis changes Others Lung Hilar Pleura score (%) Bronchitis Bronchiolitis

1 62 Restr No ev Odter fibrosis HD None Mod None 31 Insig Insig2 77 Restr No ev Silicosis Mod Mod None 0 None Insig3 62 Restr History Normal HD Slight Mod Mod 0 Insig Insig4 53 Restr No ev Silicosis HD, obese Insig Slight Insig 4 Mod Mod5 73 Obstr No ev Pleural Insig Insig Insig 9 Mod Mod6 74 Obstr No ev Normal Obese None Slight Insig 4 Insig Insig7 49 Obstr No ev Silicosis HD Mod Slight Slight 21 None None8 64 Obstr No ev Normal HD, obese Insig Mod Insig 6 Insig Insig9 86 Restr Mild Silicosis Marked Mod None 19 None None10 63 Obstr Mild Normal Obese None Slight None 23 None Insig11 66 Obstr No ev 0th path HBP Insig Slight None 0 Insig None12 50 Obstr COPD Normal HD, obese None Slight Insig 3 Insig Mod13 70 Obstr No ev Normal HD, obese Insig Slight None 17 None None14 56 Restr History Silicosis HD,HBP Marked Mod Marked 6 None None15 83 Obstr No ev Normal Obese None Insig None 17 None Insig16 78 Obstr Mild Normal HBP None None None 8 Mod Insig17 70 Obstr Mild Normal HD, obese None Slight None 19 Insig None

Restr = restrictive; obstr = obstructive; no ev = no evidence; oth path = other pathology; pleural = pleural silicosis; HD = heart disease; HBP = high bloodpressure; insig = insignificant; mod = moderate.

561

on Septem

ber 2, 2021 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.51.8.557 on 1 August 1994. D

ownloaded from

Hnizdo, Sluis-Gremer, Baskind, Murray

score was insignificant and the diagnosis ofmucus gland hyperplasia and goblet cell meta-plasia did not correspond with a history ofbronchitis in life. The medical records wereconsistent with mild bronchitis in those diag-nosed to have bronchitis in life. Thus theseresults indicate that silicosis, heart disease,and obesity are pathological factors that mayplay a part in the impairment of lung functionin this group of miners. Whether small airwaydisease is a factor in the loss of lung function inthis group cannot be resolved by this study.

It has been estimated that a career goldminer who never smokes loses on averagearound 250 ml or more of FEVI8 by 50 years ofage due to the effect of silica dust,48 and thusthere should be some significant morphologi-cal changes in the lung due to the effect ofinhalation of silica dust. The factors selectedas being negatively related to the lung func-tion tests were age, mucus gland hyperplasia,silicosis of the lung, and BMI (table 4). Thedegree of emphysema was not selected. Theaverage BMI in miners in our study was 28-8.Obesity was shown in other studies to be asso-ciated with a peripheral airways disease."Silicosis was a significant predictor of FEV,and FEV,/FVC ratio, which suggests that sili-cosis is associated with an obstructive pattern,but according to the diagnosis on the lungfunction tests of the 17 miners shown in table5, a restrictive pattern also occurs in silicoticworkers.The association between silicosis and loss

of FEVy was seen in another group of whitegold miners who never smoked and in blackgold miners who had low tobacco consump-tion.48 On the other hand, a case-controlstudy, specifically designed to assess the effectof silicosis on lung function and using 61 goldminers with radiological silicosis and 61 con-trols matched for age, dust exposure andsmoking habits did not find significant differ-ences between the two groups for a full rangeof lung function tests.'2 The exceptions werethe slope of the alveolar plateau (phase 3) andthe closing volume, for which the silicoticworkers had significantly higher values. Inthat study only eight pairs were non-smokers.

It is not certain how parenchymal silicosiscould increase airflow obstruction. Asparenchymal silicosis and emphysema wereunrelated, silicotic nodules are unlikely to actas a surrogate for emphysema. It is possible,however, that silicotic nodules act as a surro-gate for fibrosis of airways walls as in otherstudies, fibrosis of airway walls was a patho-logical feature associated with exposure to sil-ica dust.2 As radiological silicosis does notseem to be a predictor of loss of lung functionin smokers,8 this effect is likely to be smallerthan the effect of smoking or the synergisticeffect of dust and smoking on airflow limita-tion, which are known to be associated withemphysema.6An additional purpose of the study was to

establish whether the diagnosis for bronchitis,based on mucus gland hyperplasia, and ofbronchiolitis, based on goblet cell metaplasia,

as done at necropsy on the gold miners atNCOH, reflect on impairment of lung func-tion associated with exposure to silica dust.The results from this study provide no evi-dence for this. Although the degree of mucusgland hyperplasia was associated with a signif-icant reduction in FEV1, FVC, and FEV1%, itwas itself negatively associated with the degreeof silicosis of the lung (p = 0-005), and posi-tively with age at death and with BMI, andwas not associated with dust exposure. Thusfor an individual case, a positive diagnosis ofmucus gland hyperplasia or goblet cell meta-plasia at necropsy did not seem to be a reliableindicator of impairment of lung function or ahistory of chronic bronchitis, as many minerswith a moderate degree of gland hyperplasiahad normal lung function tests and did nothave a history of chronic bronchitis.The quality of the histological material for

the diagnosis of bronchiolitis, with goblet cellmetaplasia as a criterion, was mostly poor orunreadable in a sample of 24 miners selectedfor a repeatability study. It is therefore ques-tionable whether the necropsy diagnosis ofbronchiolitis based on this criteria is a correctand reliable criterion for compensation forCOPD as it is used for South African goldminers.The limiting factors in the present study

were the quality of smoking data for selectionof a sufficient number of non-smokers, thequality of the histological material for thediagnosis of the peripheral airway disease, thefact that not all the miners had lung functiontests done, the confounding effect of heartdisease and obesity, and the fact that thedegree of in life compensation reflects to someextent on the necropsy rate of the gold min-ers. The quality of the material reflects thedifficulty of the subject being studied.

Accepting these limitations, the study indi-cates, both directly and indirectly, that theimportant preventive measures for respiratoryhealth of gold miners are low respirable dustlevels for prevention of silicosis and reductionof tobacco smoking for prevention of emphy-sema.

We thank Dr Goldstein from NCOH for evaluation of the his-tological slides for bronchitis and bronchiolitis and ColleenSummerson for extracting and preparing the data.

1 Churg A, Hobson J, Wright J. Functional and morphologi-cal comparison in silica- and elastase-induced airflowobstruction. Exp Lung Res 1989;15:813-22.

2 Churg A, Wright JL, Wiggs B, Pare PD, Lazar N. Smallairways disease and mineral dust exposure. Am RevRespirDis 1985;131:139-43.

3 Wiles FJ, Faure MH. Chronic obstructive lung disease ingoldminers. In: Walton WH, ed. Inhakdparticks IV, part2. Oxford: Pergamon Press, 1977:727-35.

4 Cowie RL, Mabena SK. Silicosis, chronic airflow limita-tion, and chronic bronchitis in South African gold miners.Am Rev Respir Dis 1991;143:80-4.

5 Becklake MR, Irwig L, Kielkowski D, Webster I, De BeerM, Landau S. The predictors of emphysema in SouthAfrican goldminers. Am Rev Respir Dis 1987;135:1234-41.

6 Hnizdo E, Sluis-Cremer GK, Abramowitz JA. Emphysematype in relation to silica dust exposure, in South Africangoldminers. Am Rev RespirDis 1991;143:1241-7.

7 American Thoracic Society. Standards for the diagnosis

562

on Septem

ber 2, 2021 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.51.8.557 on 1 August 1994. D

ownloaded from

Emphysema and ainvay obstruction in non-smoking South African gold miners with long exposure to silica dust

and care of patients with chronic obstructive pulmonarydisease (COPD) and asthma. Am Rev Respir Dis 1987;136:225-43.

8 Hnizdo E. Loss of lung function associated with silica dustexposure and with tobacco smoking in South Africanwhite goldminers. BrJ IndMed 1992;49:472-7.

9 Beadle DG. The relationship between the amount of dustbreathed and the development of radiological signs ofsilicosis: an epidemiological study of South African goldminers. In: Walton WH, ed. Inhaled parties III. Oxford:

Pergamon Press, 1971:953-64.10 Hnizdo Eva, Sluis-Cremer GK. Risk of silicosis in a cohort

of white South African gold miners. Am J Ind Med1993;24:447-57.

11 Weiss ST. Pulmonary function as a phenotype physiologicalmarker of cardiovascular morbidity and mortality [edito-rial]. Chest 1991;99:265-6.

12 Wiles FJ, Baskind E, Hessel PA, Bezuidenhout B, HnizdoE. Lung function in silicosis. Int Arch Occup EnvironHealth 1992;63:387-91.

Vancouver styleAll manuscripts submitted to Occup EnvironMed should conform to the uniformrequirements for manuscripts submitted tobiomedical journals (known as theVancouver style.)

Occup Environ Med, together with manyother international biomedical journals, hasagreed to accept articles prepared in accor-dance with the Vancouver style. The style(described in full in the BMJ, 24 February1979, p 532) is intended to standardiserequirements for authors.

References should be numbered consec-utively in the order in which they are firstmentioned in the text by Arabic numeralsabove the line on each occasion the refer-ence is cited (Manson' confirmed otherreports2-5 . . .). In future references topapers submitted to Occup Environ Med

should include: the names of all authors ifthere are seven or less or, if there are more,the first six followed by et al; the title ofjournal articles or book chapters; the titlesof journals abbreviated according to thestyle of Index Medicus; and the first and finalpage numbers of the article or chapter.Titles not in Index Medicus should be givenin full.

Examples of common forms of refer-ences are:

1 International Steering Committee of Medical Editors,Uniform requirements for manuscripts submitted tobiomedical journals. BrMedJ 1979;1:532-5.

2 Soter NA, Wasserman SI, Austen KF. Cold urticaria:release into the circulation of histamine and eosino-phil chemotactic factor of anaphylaxis during coldchallenge. N EnglJ Med 1976;294:687-90.

3 Weinstein L, Swartz MN. Pathogenic properties ofinvading micro-organisms. In: Sodeman WA Jr,Sodeman WA, eds. Pathologic physiology, mechanismsof disease. Philadelphia: W B Saunders, 1974:457-72.

563 on S

eptember 2, 2021 by guest. P

rotected by copyright.http://oem

.bmj.com

/O

ccup Environ M

ed: first published as 10.1136/oem.51.8.557 on 1 A

ugust 1994. Dow

nloaded from