Document JrLV1a7EY4jVLvpkXOon9EjQO

L Ii e iiV The New England j i Journal oi Medicine I Copyright. 1971, hy the Massachusetts Medical Society Volume 285 DECEMBER 2, 1971 Number 23 EFFECTS OF LOW CONCENTRATIONS OF ASBESTOS Clinical, Environmental, Radiologic and Epidemiologic Observations in Shipyard Pipe Coverers and Controls Raymond L. H. Murphy, Jr., M.D., M.P.H., Sc.D., Benjamin G. Ferris, Jr., M.D., William A. Burgess, M.S., Jane Worcester, Dr. P.H., and Edward A. Gaensler, M.D. Abstract Low-level asbestos exposure, as occurs ! in pipe coverers in new ship construction, has not S been considered dangerous. We surveyed 101 such 1 workers and 94 controls matched for age, duration Iof employment and smoking habits in a New Eng land yard. Dust exposure, under surveillance for 20 years, had been near the then recommended threshold limit value of 5 million particles per cubic foot (mppcf). "Asbestosis" was defined by presence of three or more of five standardized clinical abnor- malities including dyspnea, rales, finger clubbing, reduced vital capacity and x-ray shadows. The lat ter were judged by three observers unaware of the exposure history. "Asbestosis" was ft times more common among pipe coverers than among controls; it was found first after 13 years of exposure or 60 mppefyears, and the prevalence was 38 per cent after 20 years. This study emphasizes that low concentrations of asbestos can lead to pulmonary fibrosis and sup ports the need for lower threshold limit values. NEW uses for asbestos are found constantly, and the number of fresh cases of asbestosis reported each year is increasing. It has been known four decades that high concentrations of dust can cause interstitial fibrosis. The biologic effects of low concentrations are still poorly defined, and informa tion useful for setting threshold limit values has been difficult to obtain. This is, no doubt, because of the many factors that must be considered to eval uate the effects of inspired dust. These include the degree of environmental contamination over a pro longed period, availability of the entire exposed population, a control group comparable in all re spects except for dust exposure and standardization of diagnostic criteria. We were able to take these aspects into account during a study of a group of shipyard pipe coverers. The concentrations of as bestos to which such workers are exposed are low1-2 except during refitting and repair work.5 4 Our workers had been engaged solely in new ship construction. From the departments of Physiology. Industrial Hygiene and Biostatitties* Harvard School of Public Health, and the departments of Med icine and Surgery, Thoracic Services. Boston University School of Medicine (address reprint requests to Dr. Murphy at the Department of Physiology. Harvard School of Public Health. 665 Huntington Ave.. Boston. Mass. 021151. Supported in part by a training grant <ES 00044). a research grant (EC 00205). a program project grant (ES 00002) and a research career award 'HE-1173) from the National Institutes of Health. U.S. Public Health Service. The data contained in this study are from a thesis submitted by Dr. Murph) to the Faculty of the Harvard School of Public Health in par tial fulfillment of the requirements for the degree of Doctor of Science in the field of Environmental Health. We wished to determine the effect of this lowlevel exposure to ascertain whether the thresh foor ld limit value had been a reasonable guideline. In addition, we were interested in evaluating methods to detect the disease at the earliest possible stage to prevent further exposure. Environmental, clinical and epidemiologic aspects of this survey will be presented here, and studies of respiratory gas ex change will be reported subsequently. Environmental Survey The pipe coverers were employed in a private New England shipyard that had been engaged in construction of naval arid commercial vessels since 1920. During World War II, it became principally a Abbreviations Used O.: FEV,: FVC: mppcf: PF: diffusing capacity volume expired in first second forced vital capacity million particles per cubic foot peak flow destroyer-building facility. Repair and refitting work was never done. The yard was one of five investi gated by Fleischer et al* in 1945. The pipe coverers prepared and applied insulating materials to machinery and pipes. The materials used and the procedures employed did not van- appreciably over the 20-year period.* Approximately 90 per cent of the work was done aboard ship and the remainder in a shop. PLAINTIFF'S exhibit AL-969 i 1272 THE NEW ENGLAND JOURNAL OF MEDICINE Dec. 2, 1971 Figure 1. Hand-Sawing of Prefabricated Asbestos Block Aboard Ship. Aboard ship the most common procedure was wiring contoured asbestos block to pipes. Prefabri cated sections were tailored to fit bends by handsaw and knife (Fig. 1). The blocks were smoothed with cement containing 15 per cent asbestos and 85 per cent magnesia, and covered with asbestos cloth. Asbestos in fibrous form was stuffed into jackets and sewn around small pipes. Ventilation was provided by portable blowers on deck, but they were not always used. Shop work included layout, cutting, sewing, as bestos cement mixing and fabrication of pads. In 1963, a ventilation exhaust trunk was installed over the four benches. Band-saw cutting, cement mixing and pattern cutting were dustier procedures but required less than 5 per cent of man-hours, and involved intermittent work by only a few men. Pat tern cutting was done in a separate room by one man who wore a respirator. The insulating materials were largely determined by United States Xavy specifications. Amosite was called for most frequently because of ite low ther mal conductivity, light weight and strength. Chrysotile was used to a lesser extent, and crocidolite was never used. Fiberglass, cotton, calcium silicate, magnesia, hair felt and other materials were also employed, either alone or mixed with asbestos. Aerometric Analysis Dust concentrations were determined in 1945* and again in 1965 and 1966, to coincide with these studies. The same konimeter was used according to the method of Kotze. All dust particles in a highpower field were counted for the total dust count. Particles with a length-to-diameter ratio greater than 3 were called fibers. Both dust and fiber concentra tions were ex-pressed in millions of particles per cubic foot of air (mppcf). Simultaneous total dust counts were made with a midget impinger in 1965 according to the Public Health Service method. All counts were made at the worker's breathing level at all the usual sites during usual operating conditions. A time-weighted average dust exposure was calculated by multiplying the mean dust con centration at various sites by the average percentage of time spent by the workers at these sites. This was expressed in mppef-years. Results of dust counts with the midget impinger -- the instrument used in setting the threshold limit value of 5 mppcf -- averaged 5.2 mppcf (Table 1). Konimeter counts, as always, were higher: in areas where over 98 per cent of the man-hours were spent the overall range was from 8.2 to 61.7 mppcf (Table 1). A weighted average, which considered the amount of time spent in these locations, as well as the number of samples taken during the three periods in 1945, 1965 and 1966, was 32.7 mppcf. Population Studies There were 101 pipe coverers employed at the Table 1. Dust Concentrations in Million Particles per Cubic Foot. A. Midget Impinger (1965V*: LOCATION Curling room Sewing & fabrication bench Band-saw cutting Mixing mud Aboard ship (various) Weighted average Konimeter location VR Sewing & fabrication Aboard ship 1945 1965 1966 1945 1965 1966 'Instrument used in setting threshold limit value of 3 mppcf. AVERAGE CONCENTRATION (MPPCF) 5.7 3.0 10.0 0.8 7.2 5.2t - NO. OP SAMPLES 2 12 10 15 12 10 CONCENTRATIONS (MPPCF) rouge average 10.6-12.3 14.2-30.2 12.9-32.8 25.3-89.0 8.2-61.7 18.3-35.7 11.4 23.4 23.1 49.2 21.4 25.9 % FIBER 0.5-0.8 0.1-2.7 0.1-3.8 0.2-2.4 0.1-0.7 0.1-0.7 tSimulianeous counts uith Konimeter equaled '4.4 mppcf. Vd. 285 No. 23 LOW CONCENTRATIONS OF ASBESTOS-MURPHY ET AL. 1273 Yard- in November, 1965. One hundred and one controls were selected from the 486 shipfitters and pipefitters employed at die time of the survey. To match them for age and duration of employment, cards with this information only were inspected in sequence, and the first worker whose age was with in one year of a pipe coverer was selected provided he had spent approximately the same time at the Yard.' All pipe coverers participated, but seven of die selected controls refused, leaving 94 individ uals. Methods Questionnaire A questionnaire on respiratory symptoms7 was administered by one observer. It was adapted from that previously given extensive trials by Fletcher and others.*'10 Past History The company medical records of all workers were examined for symptoms, signs and diagnostic im pressions presumably related to cardiorespiratory disease. Physical Examination Physical examination was made by one physician with particular attention to breath sounds in each of eight locations. In 1966, tape recordings were made. Adventitious sounds were defined according to the nomenclature of Chamberlain.11 To evaluate club bing, a tracing of all fingers was made in the frontal and lateral projection. The hyponychial angle of the right index finger was measured according to Regan et al.,11 and an angle of 198 or greater was consid ered indicative of clubbing. Spirometry Maximal forced expirations were recorded with a Stead-Wells spirometer, and peak flow with a Wright Peak Flow Meter. The order of tests was randomized. Five recordings were made, and the average of the last three was used to calculate the forced vital capacity (FVC), the volume expired in the first second (FEV,) and the peak flow (PF). Methods of calibration and analysis were previously described'1; normal predicted values derived from a population study of a New England town were used.14 Roentgenologic Examination Posteroanterior and lateral views were exam ined independently by three radiologists after identification had been masked and the films had been randomly interspersed. Results were coded as follows: 1, no abnormality; 2, abnormality not con sistent with asbestosis;' 3, questionably consistent with asbestosis; 4, consistent with slight asbestosis; 5, consistent with moderately advanced asbestosis; and 6, consistent with advanced asbestosis.* The reported results represent the combined readings of the three observers. Prevalence of Other Disease The prevalence of tuberculosis was assessed from the roentgenograms, skin testing and sputum cul ture. The criteria of Fletcher and Tinker1* were used to calculate the prevalence of nonspecific chronic obstructive lung disease. Simple and com plex bronchitis was defined according to Reid et al.17 Results Description of the Populations The control group, matched for age and duration of employment, was similar to the exposed group in color, height, weight, marital status and location of residence (Table 2). The majority of both groups had spent their lives in the general vicinity' of the shipyard town. All were blue-collar workers who received hourly wages. The pipe coverers had been employed at the Yard for an average of 17.4 years, and the controls for 17.1 years. With shipbuilding for defense, the work force var ies with national crises. When not working at the Yard, members of both groups most commonly en gaged in fishing, lumbering and general outdoor construction work. Cumulative years in dusty occu pations other than pipe covering did not differ in the two groups. One pipe coverer had been a sandhog for five years, and one control had spent five years shoveling gravel. There were no exposures to material likely to cause pneumoconiosis in less than five years of exposure. Respiratory Questionnaire The significant differences in positive responses appear in Table 2. Smoking habits were similar: according to our classification,1* two thirds of both groups were "present smokers" (Table 2). Physical Examination Dry, crackling rales with a "close-to-the-ear" sound (cellophane rales) were a most striking finding in some pipe coverers. They could be clear ly identified over background noise in the tape re cordings, whereas occasional fine rales could not be distinguished. Such rales not only were significantly more common in the pipe coverers (Table 2) but also were heard more often in multiple sites. For example, they occurred in three or more sites in seven pipe coverers but never in the controls. Four years afler these interpretations were made, a modification of the international Scheme for Classification of the Radiographic Ap pearance of Pneumoconioses was proposed that was believed to be suitable for interpretation of asbestosis.'* Our classification corre sponds approximately as follows: "I", 0/0: "3". 0/1: I/O. I/I and 1/2: "5". 2/1. 2/2 and 2/3: and "6". 3/2. 3/3 and 3/4. 1274 THE NEW ENGLAND JOURNAL OF MEDICINE Dec. 2, 1971 Table 2. Summary of Selected Results. Finding Age(yr) Height (m) Weight (lb) respiratory SYMPTOMS Cough -- day -- winter Cough -- day -- summer Cough 3 yr Phlegm - day -- winter Phlegm - day - summer Wheezing apart from colds Breathlessness (Fletcher 111 or more) Shortness of breath. 1 flight `Heart trouble'* or "high blood pressure" SMOKING CATEGORY Never smoked Ex-smoker Presenr smoker Uncertain recording PHYSICAL EXAMINATION Basilar rales: Right lower lobe Left lou er lobe 2 or more sites Clubbed fingers - hyponvchial angle >19S* PULMONARY FUNCTION FVC (liters) FEV, (liters) PF (liters/min) FVC 5? of predicted ROENTGENOLOGIC PANEL INTERPRETATION (1) Norma) (2) Other abnormality (3) Questionable asbestosis (4) Slight asbestosis (5) Moderately advanced asbestosis (6) Advanced asbestosis tot Pipe Coverers MUK SD 41.5 67.9 168.0 11.6 2.3 23.5 NO. POSITIVE 51 21 25 39 34 29 38 26 20 rA POSITIVE 30.7 20.8 24.8 38.6 33.7 28.7 37.6 25.7 19.8 NO. POSITIVE 7 27 67 0 no. posmvt 91 POSITIVE 6.9 26.8 66.4 0 POSITIVE 94 Controls MEAN SO 40.9 68.4 174.7 ) 1.6 2.7 27.5 NO. POSITIVE 9E POSITIVE 18 19.1 9 9.6 9 9.6 23 24.5 12 12.8 15 16.0 23 24.5 6 6.4 15 16.0 NO. POSITIVE 7 23 62 2 9 POSITIVE 7.4 24.5 66.0 2.1 NO. POSITIVE 9E POSITIVE 23 16 16 20 MEAN 3.9 3.0 478.8 92.7 SO. POSITIVE 28 2 27 31 9 4 22.8 15.8 15.8 19.8 , to 0.76 0.77 97.9 1.3 "X POSITIVE 27.8 1.9 26.7 30.7 8.9 4.0 5 7 4 5 MEAN 4.3 3.6 526.3 102.7 NO. POSITIVE 52 8 13 19 2 .0 5.3 7.4 4.2 5.3 so 0.87 0.74 84.6 1.7 5 POSITIVE 55.5 8.5 13.8 20.1 2.1 0 p Value NS NS NS NS <0.005 <0.01 <0.05 <0.005 <0.01 <0.05 <0.005 NS NS NS NS NS <0.005 NS <0.01 <0.005 <0.001 <0.001 <0.001 <0.001 <0.001 Wheezes, rhonchi and moist rales were present equally often in both groups. Ventilatory Tests There was a marked difference in mean FVC . (Table 2). Peak flow and FEV, were also 'significantly lower in the exposed workers, and were close to the predicted level in the control group (Table 2). Roentgenologic Findings Moderately advanced (code 5) or advanced asbestosis (code 6) was diagnosed six times more fre quently in pipe coverers than in the controls, whereas slight asbestosis (code 4) occurred only 1.4 times more frequently (Table 2). For each of the three observers, the chances were good that the classification agreed with the exposure history when the readings were moderately advanced or advanced; the chances were much smaller when the classifi cation was questionable or slight asbestosis. Of the two controls with a reading of "moderately advanced asbestosis," one had previously recognized pulmonary fibrosis after smoke inhalation 30 years ago, and the other had bullous emphysema with diffuse linear opacifications. Medical-Record Review Forty-seven pipe coverers and 35 controls had notations concerning cardiopulmonary complaints, an insignificant difference (p greater than 0.05). However, pleurisy, pleura] effusion and pleural rub were noted 13 times in pipe coverers and four times in controls. Other notations that were more common in pipe coverers included bronchitis (10 and three), influenza (15 and four) and dyspnea (five and zero). Prevalence of Asbestosla To assess the prevalence of this disease it became necessary to define criteria. Manifestations most commonly reported in persons with known asbesto sis were used1-15: dyspnea on climbing one flight of stairs or less; basilar rales in two or more sites; clubbing of the fingers -- that is. a hyponvchial angle of 198 or greater; a vital capacity- of less than 80 per cent of the predicted value; and a roentgeno gram consistent with moderately advanced or ad- Vol. 285 No. 23 LOW CONCENTRATIONS OF ASBESTOS-MURPHY ETAL. 1275 vanced asbestosis (codes 5 or 6). The presence of at least three of these signs was defined arbitrarily as necessary for an epidemiologic diagnosis. By this definition, 11 pipe coverers had "asbestosis"; all 11 had roentgenograms coded 5 or 6, nine had reduced vital capacities, nine clubbing, eight basilar rales, and eight exertional dyspnea. Also, by this definition, one of the controls had "asbestosis" -- the worker who had pulmonary fibrosis after smoke inhalation. Respiratory Gas Exchange In subsequent surveys, one and three years later, we studied the role of several physiologic tests in defining and detecting this disease. Pipe coverers had significant!)' reduced single-breath and exercise steady-state diffusing capacities (DJ whereas air ways resistance, specific conductance, ventilation, carbon dioxide tension and dead space were not significantly different. Physiologic evidence of ob structive disease was equally common in both groups. All workers with clinical "asbestosis" had severely reduced DL, corroborating our clinical cri teria (Murphy, R. L. H., Jr., Caensler, E. A., Red ding, R. A., et al: Unpublished data). Figure 2 shows the relation of data on respiratory gas exchange to our clinical findings. Our diagnostic criteria that focused on marked abnormalities corre lated highly with impaired DL. However, Figure 2 also suggests that clinical findings of lesser severity' may reflect interstitial lung disease. Prevalence ot Other Respiratory Diseases Skin hypersensitivity to 5 tuberculin U of PPD was not significantly different in the two groups. All persons with 5 mm or more of induration submitted X-ftoy Clubbing Dry RoltS Oyspnto Vita Copoe % Sit 314 t >97 B-9T <9 2 I 0 2 I 0 *10 P-40*90 i4 a /a t it /a av ha Figure 2. Relation between Du (Steady State at Rest, DLSS, R; Exercise. DLSS. Ex; Single Breath, DLSB) and the De gree of Five Clinical Abnormalities among Pipe Coverers and Controls (Figures at Bottom Represent Numbers of Sub jects with the Indicated Abnormality). Table 3. Chronic Nonspecific Obstructive Respiratory Dis ease. Category 1 -- no obstructive lung disease 2 -- chronic bronchitis only* 3 -- asthma onlyt 4 -- chronic obstructive lung disease onlyt 101 Pin Coverers tt 71.3 18.8 0 3.0 94 Shipyard Controls % 68.) 25.5 0 3.2 Combinations 2+3 3+4 2+4 2+3 + 4 1.0 0 5.9 2.1 00 0 1.1 Phkgm from chesi at kail 6 timei/day. 4 days/k for at least 3 mo for past 3 yr. tAsthma diagnosed by a physician & still present. |Wheezing or w hisiling most days or nights, or worker has to stop for breath w hen walking at own pace on the level, or FEV, <605 of F\'C. three 24-hour sputum specimens, none of which were positive for Mycobacterium tuberculosis. Nonspecific obstructive lung disease was slightly but not significantly more common in the control group and among cigarette smokers (Table 3). Among workers with this syndrome, approximately 75 per cent of the pipe coverers and 65 per cent of the controls smoked cigarettes. Asbestosis and Its Relation to Dust Exposure By our scheme, "asbestosis" was never diagnosed in persons with less than 10 years of cumulative exposure, with a progressive increase thereafter (Fig. 3). If uniform exposure over the working ca reers is assumed, certain conclusions can be reached from dust counts made with the midget impinger. No "asbestosis," as defined above, oc curred in men exposed for less than 60 mppcf-years; 20 per cent of those exposed for 75 to 100 mppcfyears were considered to have asbestosis, and the prevalence was 38 per cent in those with exposure for more than 100 mppcf-years. Discussion Building-insulation workers move from site to site and are confronted by varying situations and specifications, making supervision and assessment of dust exposure difficult. An important study confined to such workers by Selikoff, Churg and Hammond1 showed that almost Vi of 1117 workers were con sidered to have roentgenologicallv evident asbesto sis; prevalence was directly related to duration of exposure. By contrast, shipboard pipe coverers en gaged in new vessel construction have a more uniform exposure. Fleischer et al.,* in a studs- al ready alluded to, found only three cases of asbesto sis among 1074 pipe coverers, and these three had been exposed for 20 or more years. Therefore, the authors thought that such pipe covering was not a 1276 THE NEW ENGLAND JOURNAL OF MEDICINE Dec. 2, 1971 Figure 3. Relation of "Asbestosis" to Duration of Expo sure. "Asbestosis" was never diagnosed in persons with less than 10 years of cumulative exposure, and the prevalence was 38 per cent in those exposed for more than 20 years. One member of the control group had "asbestosis." dangerous occupation. More recently, Marr* reported five cases of disability among 60 to 80 shipyard pipe coverers; he questioned whether this was caused by massive exposure during removal of old insulation or from many years of exposure by sus ceptible persons during all types of insulation work. Our principal objective was to determine the prevalence of asbestosis in pipe coverers who had not engaged in repair work. Diagnostic criteria were based on recognized symptoms and signs of this disease because an exact assessment would have required pathological material. These signs and symptoms have been reported in other defined populations of asbestos workers with similar frequency.,*,,',,,4 The prevalence of shortness of breath among our pipe coverers (Table 2) was similar to the 12 to 26 per cent in otheT asbestos workers.In unex posed North American populations this percentage has ranged between 2.6 and 5.77',,J5-M -- quite comparable to our own control group with 6.4 per cent Rales in two or more sites were found in 15.8 per cent of the pipe coverers -- a rate not very different from the 13 to 14 per cent reported by others in asbestos workers.1'41 In our study the observer was unaware of the roentgenographic findings at the time of ausculta tion. Nevertheless, 33 per cent of workers with codes 5 and 6 roentgenograms had rales in two or more sites. Dreessen et al.11 reported rales or other adventitious sounds in 49 per cent of those with "ground-glass" markings, whereas Vigliani" heard such sounds in 70 per cent of workers with abnor mal roentgenograms. Clubbing of the fingers in 19.8 per cent of pipe coverers was more frequent than that in other studies11'*4; also, the 5.3 per cent prevalence in die controls was slightly larger than in otheT normal groups.11 However, the mean hyponychial angle of controls considered clubbed was only- 200 as com pared to that of the clubbed pipe coverers, which was 208*. Vital-capacity measurements were usually lacking in the earlier surveys.'4141 Kleinfeld14 reported an average FVC of 81 per cent of predicted in 56 workers exposed for more than 14 years, but he used different prediction formulas. The mean FVC of 4.0 liters in 21 Australian asbestos workers did not differ much from that of our pipe coverers. Hunt's exposed population1* is difficult to compare because knowledge of the diffusing capacity is re quired for interpretation. Our regression equations were derived from a population survey.14 Since such surveys include disabled people, they generally yield lower values than surveys of industrial work ers. For example, our pipe coverers had an average FVC of 92.7 per cent based upon our population survey, whereas it was only 75.2 per cent with the use of normal values from 2770 persons engaged in heavy labor and mining. Pipe coverers over the age of 55 had a mean FVC that placed them in the same category as workers with advanced silicosis.1* The FVC of all pipe coverers was also lower than the mean of paper-mill, flax and polyurethane-foam workers.**4*1** Concerning radiographic findings, some observers have considered early changes evidence of asbesto sis,14,41 whereas others have not.11'1*'**'*4 At any rate, our 19.9 per cent prevalence of moderately ad vanced or advanced asbestosis (codes 5 and 6) was much higheT than the 0.27 per cent in the only other survey of shipyard pipe coverers engaged in new vessel construction,1 but it was about the same as in building-insulation workers.1 Effect of Duration of Exposure and Age The prevalences of clinical signs and symptoms of "asbestosis" increased with duration of exposure (Fig. 4). The prevalences also rose with age (Fig. 5), and therefore the association between age and cumulative years of exposure was high (r equal to 0.7, p equal to 0.001). The advantage unique to this study was that the control group allowed examina tion of the relative importance of age: in the con trols there was no clear association between age and these abnormalities except in vital capacity (Fig. 5). Accuracy of Diagnosis How likely is it that the 11 workers considered to have "asbestosis" did indeed have this disease? They were offered a medical examination, including Vol. 285 No. 23 LOW CONCENTRATIONS OF ASBESTOS-MURPHY ET AL. Pipe Cov Controls II 20 10 21 26 - 29 21 32 19 10 1277 Exposure. Years * Oyspnta -- * Rales ----- % Clubbing --* YC.<BO --X-Roj 596 0 1 Figure 4. Increases with Duration of Exposure in Preva lence of Dyspnea, Rales, Clubbed Fingers, Reduced Vital Capacity and Roentgenologic Abnormalities. detailed pulmonary-function studies, by a group of internists not associated with this study. In the eight who consented, no illness other than asbestosis was discovered to account for the signs and symptoms. Two of these men have since died, and both had severe asbestosis. Of the remaining three who refused detailed examination, two have died; autopsy in one showed asbestosis and cor pulmo nale, and the other died of bronchopneumonia. The accuracy of a "negative" diagnosis is much more difficult to assess. Only one worker so classified has died. His history, physical examina tion and ventilatory-function studies had all been normal at the time of this survey, and his chest roentgenogram showed slight asbestosis (code 4). When he died of peritoneal mesothelioma, the lungs showed minimal chronic interstitial pneumo nitis. Conclusions should not be drawn from this single case, but the prev alences reported here proba bly are too low. Our subsequent surveys were de signed to investigate this possibility. What is the likelihood that other disease caused the findings that we attributed to asbestosis? Medical-record review, the respiratory' questionnaire, physical examination, skin testing, sputum examina tion and roentgenograms did not uncover any excess cardiac disease, tuberculosis or other pulmonary disease. There was no significant difference in smoking habits in the entire population of pipe coverers and controls. Asbestosis and Dust Exposure TIk- concept of a dose-response relation was intro duced by Merewether and Price in 1930,'* but the first information useful for calculating threshold lim- ---- -101 Pip* Cwenra --94 Control* . Figure 5. Higher Prevalences of Clinical Abnormalities Seen in Older Pipe Coverers but Not in Older Controls. Since the association between age and cumulative years of exposure (Fig. 4) was high (r equal to 0.7, p less than 0.001). the importance of aging in relation to these abnor malities was reviewed in the controls; no clear association with age was found except in vital capacity. it values came from Dreessen et al. in 1938.*' They found clear-cut asbestosis among those exposed to dust concentrations exceeding 5 mppcf, and none at lower concentrations. Therefore, tentatively, they regarded 5 mppcf as a threshold value.11 These ob servations were confirmed by Vigliani.11 In retro spect, the choice of 5 mppcf, on the basis of the data then available, was open to question; in the dust counts in the textile mills no distinction was made between cotton and asbestos fibers. Further more, among those who were exposed at or just above this threshold (that is, 5.0 to 9.9 mppcf), 13 of 19 were considered to have asbestosis after 10 years, and six of 37 after only five years. In relation to the dose-times-time relation, over V2 the workers exposed to 100 mppef-vears had asbestosis. Divid ing 100 mppef-years by the recommended threshold 1278 THE NEW ENGLAND JOURNAL OF MEDICINE Dec. 2, 1971 limit value of 5 mppcf should have prompted the conclusion that the risk of asbestosis was great after only 20 years of exposure. Indeed, other observa tions have suggested that asbestosis becomes evi dent after approximately this degree of total dust exposure: Wells** began to see the disease after 50 to 60 mppcf-years, and Fulton** reported a 17 per cent prevalence after exposure from 27 to 75 mppcfyears. In our study, asbestosis became evident after a total exposure of 60 mppcf-years, and the earliest case was found after 13 years of employment Fleischer's prevalence in the same shipyard was much lower because, in 1945, less than 10 per cent of the workers had been exposed for more than 10 years. These observations indicate that prolonged exposure to low concentrations of asbestos is haz ardous, and support the lowering of the thresholdlimit values. We are indebted to Drs. Kenneth T. Bird, Eugene P. Pendergrass and Alfred L. Weber, who read, graded and classified the roentgenograms, and to Mrs. Lee Slocum, who wrote the computer programs. References 5. Selikoff IJ. Churg J, Hammond EC: The occurrence of asbestosis among insulation workers in the United States. Ann NY Acad Sci 132:139-153, 1965 2. Fleischer WE. Viles FJ Jr. Gade RL, et al: A health survey of pipe covering operations in constructing naval vessels. J Ind Hyg Toxicol 28:9-16. 1946 3. Marr WT: Asbestos exposure during naval vessel overhaul. Am Ind Hyg Assoc J 25:264-268. 1964 4. Stumphius J. Meyer PB: Asbestos bodies and mesothelioma. Ann . Occup Hyg 11:283-293. 1968 J. Miners Phthisis Prevention Commitiee: Final report. Johannes burg. Union of South Africa, 1919 6. American Conference of Governmental Industrial Hygienists, 1962: Air sampling instruments (Document B-3-31). Cincinnati, Ohio, 1963 7. Ferris BG Jr. Anderson DO: The prevalence of chronic re spiratory disease in a New Hampshire town. Am Rev Resp Dis 86:165-177, 1962 8. Fletcher CM, Elmes PC. Fairbaim AS, et al: The significance of respiratory symptoms and the diagnosis of chronic bronchitis in a working population. Br Med J 2:257-266. 1959 9. Higgins ITT: Respiratory symptoms, bronchitis, and ventilatory capacity in random sample of an agricultural population. Br Med J 2:1198-1203. 1957 10. Fairbaim AS, Wood CH, Fletcher CM: Variability in answers to a questionnaire on respiratory symptoms. Br J Prev Soc Med 12: 175-193, 1959 11. Chamberlain EN, Ogilvie C: Symptoms and Signs in Clinical Medicine: An introduction to medical diagnosis. Eighth edition. Bristol, England, Wright and Sons, 1967 12. Regan GM. Tagg B, Thomson ML: Subjective assessment and objective measurement of finger clubbing. Lancet 1:530-532, 1967 13. Clinical spirometry: recommendations of the Section on Pulmo nary Function Testing. Committee on Pulmonary Physiology. American College of Physicians. Dis Chest 43:214-219, 1963 14. Ferris BG Jr. Anderson DO, Zickmantel R: Prediction values for screening tests of pulmonary function. Am Rev Resp Dis 91:252- 261. 1965 15. Bohlig H. Bristol L). Cartier PH. et al: UICC/Cincinnati classi fication of the radiographic appearances of pneumoconioses: a cooperative study by the UICC Committee. Chest 58:57-67, 1970 16. Fletcher CM. Tinker CM: Chronic bronchitis: a further study of simple diagnostic methods in a working population. Br Med J 1: 1491-1498. 1961 17. Reid DD. Anderson DO, Ferris BG Jr, et al: An Anglo-Ameri can comparison of the prevalence of bronchitis. Br Med J 2:1487- 1491. 1964 18. Anderson DO, Ferris BG Jr, Davies TW: The Chilliwack Re spiratory Survey. 1963. Part I. Methodology. Can Med Assoc J 92:899-905. 1965 19. Merewethcr ERA, Price CW: Report on Effects of Asbestos Dust on the Lungs and Dust Suppression in the Asbestos Industry. London, His Majesty's Stationery Office. 1930 20. Ellman P: Pulmonary asbestosis: its clinical, radiological, and pathological features, and associated risk of tuberculous infection. J Ind Hyg 15:165-183. 1933 21. Dreesscn WC. Dallavalle JM, Edwards Tl. et al: A study of as bestosis in the asbestos textile industry. Public Health Bulletin No 241. Washington. DC. Government Printing Office, 1938 22. Vigliani EC: Studio sull'asbestosi nelle manifatiure di amianlo. Cine. G. Capella, 1940 23. Selikoff 1J. Bader RA. Bader ME, et al: Asbestosis and neoplasia. Am J Med 42:487-496. 1967 24. Kleinfeld M, Messite J, Kooyman O, el al: Effect of asbestos dust inhalation on lung function. Arch Environ Health 12:741-746. 1966 1 25. Ferris BG Jr, Anderson DO, Burgess WA: Prevalence of re spiratory disease in a flax mill in the United States. Br J Ind Med 19:180-185. 1962 26. Ferris BG Jr. Burgess WA. Worcester J: Prevalence of chronic respiratory disease in a pulp mill and a paper mill in the United States. Br J Ind Med 24:26-37, 1967 27. Sayers RR. Bloomfield JJ. Dallavalle JM. et al: Anthraco-Silicosis Among Hard Coal Miners. Public Health Bulletin No 221. 1935 28. Hunt R: Routine lung function studies on 830 employees in an asbestos processing factory. Ann NT' Acad Sci 132:406-420, 1965 29. Arbeitsphvsiologische und arbeitspathologische Studien: Schriften- reihe fur Arbeitshygiene und Arbeitsmedizin. Vol 1. Luxem burg. Europaische Gemeinschaft fur Kohle und Stahl, 1961, p 103 30. Peters JM, Murphy RLH, Pagnotto LD. et al: Acute respiratory efTects in workers exposed to low levels of toluene diisocyanate (TDD. Arch Environ Health 16:642-647. 1968 31. Lanza AJ. McConnell WJ. Fehnel JW: Effects of the inhalation of asbestos dust on the lungs of asbestos workers: a preliminary study. Public Health Rep 50:1-12. 1935 32. McPheeters SB: A survey of a group of employees exposed to asbestos dust. J Ind Hyg 18:229-239. 1936 33. Cartier P: Some clinical observations of asbestosis in mine and mill workers. Arch Ind Health 11:204-207, 1955 34. Bohlig H. Jacob G. Muller H: Die Asbestose der Lungen: Genese. Klinik, Rontgenologic. Stuttgart, Germany, Georg Thieme. 1960 35. Wells J: Biological effects of asbestos. Section V. Human exposure to asbestos: dust controls and standards: discussion. Ann NY Acad Sci 132:335-336. 1965 36. Fulton WB, Dooley A. Matlhews JL. et al: Asbestosis. Part II. The nature and amount of dust encountered in asbestos fabricat ing plants. Part III. The effects of exposure to dust encountered in asbestos fabricating plants on the health of a group of workers (Bulletin 42). Pennsylvania Department of Labor and Industry. Harrisburg, 1935