Symptoms and clinical findings in patients with silicosis.

KOSKINEN H. Symptomsand clinical findings in patientswithsilicosis. Scand J Work Environ Health 11 (1985) 101-106. Respiratory symptoms, radiographic findings, and lung function were analyzedin 144 Finnishpatients with silicosis. The prevalence of persistentphlegmproduction in these patients was 46 "To and that of dyspnea 87 0/0. Impairment of the vital capacity (VC) and the diffusion capacity (DLco) « 80 070 of predicted values) were found in 46 and 47 0J0 of the patients with simplesilicosis, respectively. The mean DLco was lower in patients with advancedsimplesilicosis (category 3) than in those with slight simplesilicosis (categories I and 2). In category3 the DLco was impaired in 9 out of 12patients, the impairmentbeingbelow65 % of the predictedvalues in sixof the nine. All patients with largeopacitiesshowedimpairmentin their lung functiontests. Twenty-eight of the silicosis patients had referentsmatched for exposureto silicadust, age, and gender.The referents had no radiographic signs of silicosis. The patients experienced dyspneamore often than their referents, whereas no differencewas found in the prevalence of persistentphlegmproduction. Themean valuesof ve, forcedexpiratoryvol ume in I s (FEVl.o), and DLco were lower in the patients than in the referents. The results indicate a high prevalence of dyspnea, restrictive impairmentof lung function, and impaired diffusion capacityin the patients with simpleor complicated(ie, large radiographicopacities) silicosis. The DLco proved to be a rather sensitive lung function parameter for advanced simplesilicosis.

Silicosis may occur in occupations where siliceous materials are encountered. Patients with simple silicosis -that is, with multiple, rounded opacities of various profusion as seen in a chest radiographhave no respiratory symptoms (12). Symptoms indicative of chronic bronchitis are equally common in silicotics and in subjects similarly exposed but without silicosis (3,12). Dyspnea on effort is the most common feature of silicosis (4,14); this symptom is, however, usually associated with complicated (eg, cases with large radiographic opacities) rather than with simple silicosis (14), which does not usually produce significant impairment in lung function (4,8,12,14). The focal fibrosis may be associated with a slight decrease in vital capacity or total lung capacity (12), or it may produce spirometric evidence of small airways disease or evidence of impaired gas mixing (4).
The principal symptom of patients with complicated silicosis -that is, eg, with large radiographic opacities -is shortness of breath on exercise (4,12,14). The functional pattern in such cases may be one of obstruction, mixed obstruction and restriction, or pure restriction (4). The diffusion capacity at this stage is reduced (12) .
The objective of the present study was to assess the interrelationships between respiratory symptoms, radiographic findings, and lung function variables in a series of Finnish silicosis patients. The patients were compared with referents matched for silica dust exposure, age, and gender, but with no radiographic signs of silicosis.

Subjects
A total of 306 cases of silicosis were reported to the Finnish Occupational Disease Register in 1965-1977. The patient group for this clinical reexamination of persons with silicosis was obtained from these reported cases . The study was performed in 1979 at the Institute of Occupational Health in Helsinki. The diagnostic criteria for silicosis included a verified history of occupational exposure to silica dust, or mixed dust including silica, and a radiographic finding consistent with silicosis. To avoid false positive cases of silicosis, I reevaluated the occupational history and clinical data from the patient records of the 306 reported cases of silicosis and also reviewed the chest radiographs together with two radiologists. A total of 243 of the persons reported to have silicosis were alive, and 63 had died before 1979. All 243 living silicosis patients had a verified history of occupational silica dust exposure. Twenty-five of them were excluded from the reexamination because the radiographic or clinical findings were inconsistent with silicosis. Twenty patients could not attend the reexamination because of poor health, mainly due to old age. Forty could not be traced, and 14 refused the reexamination. Thus 144 patients (138 men and 6 women) with silicosis were available for reexamination, and they formed the patient group of my study. The data on their gender, age, occupational history, and duration of exposure to silica dust are presented in table 1.
Twenty-three (16 0,10) of the 144 silicosis patients had a history of pulmonary tuberculosis, which was active in two patients at the time of the reexamination.
In order that the effect of silica dust on symptoms and clinical findings could be studied in nondiseased workers, an attempt was made to find a referent for as many of the silicosis patients as possible. Matching one referent to one patient for gender, age (± 5 years), and duration of exposure (± 5 years) to silica dust in the same workplace and similar working tasks produced 28 male pairs. One of the referents was, however, nine years younger than his matched patient. The data on age, occupational history, and duration of silica dust exposure, as well as the data on the radiographic findings, of these 28 patients and referents are given in table 2. Five of these patients, in contrast to 25 referents, were currently exposed to silica dust when the examination was performed, whereas the other patients had been exposed 1 to 14 (mean 5) years earlier.
The smoking habits of the 28 patients and their referents, as well as of the patients in various radiographic profusion categories, were similar. Of the 144patients, 25 were nonsmokers, 71 exsmokers, and 48 current smokers.

Methods
The occupational and medical histories of the patients with silicosis and the referents were obtained in interviews carried out by myself. A questionnaire of the Medical Research Council (7) was used in the assessment of the respiratory symptoms. A smoker was defined as a subject who still smoked or who had given it up less than three months prior to the interview, an exsmoker was someone who had given up smoking more than three months before, and a nonsmoker was a person who had never smoked regularly. Persistent production of phlegm was defined as the production of phlegm for at least three months each year for two years or more. The definitions for various degrees of breathlessness are given in table 3. The radiographic examination of the patients and referents included two full-size posteroanterior and one lateral radiograph. The radiographic technique has been described in detail elsewhere (15). The chest radiographs of the patients and referents were classified independently by two physicians (one radiologist and myself) with experience in reading chest radiographs of pneumoconiosis patients using the 1980 classification of radiographs of pneumoconioses published by the International Labour Organisation (ILO) (2). The radiographs were classified without access to the identification data on the films. If the two classifiers did not agree in their independent grading assessment , the films were reviewed jointly, and a category was agreed upon.
Large opacities, as defined by the ILO classification (2), were recorded for only nine patients with silicosis, the number of patients with simple nodular silicosis therefore being 135. Six of the nine had a profusion of small opacities of category 3, and th ree a profusion of category 2.
A Bernstein type of spirometer (Kifa, Sweden) was used to measure the vital capacity (VC), the forced vital capacity (FVC) , and the forced expiratory  (11). The ventilatory impairment was considered to be obstructive if the FEV% was at least I3 percentage units less than the predicted value (FEV% < -13 % of predicted) and restrictive if the ve was less than 80 % of the predicted value. Values less than 80 % of the predicted ve , FEV I.O , and OLeo refers to values lower than the mean minus two standard deviations and those less than 65 % o f predicted to values lower than the mean minus four standard deviations of the reference values .
The statistical significance of the differences in the prevalence data on the cross-sectional unpaired samples was tested with the Pearson chi-square test; in comparisons with low frequencies, Fisher's exact probability test was used. When the cases and their referents were compared, the McNemar test or the ttest for pairwise differences was used. The Student's t-test was applied to the significance testing of differences in lung function between unpaired groups.

Symptoms of patients with silicosis
Sixty-six (46 %) of the 144 patients with silicosis had a persistent production of phlegm; 10 (15 %) of the 66 were nonsmokers. Persistent phlegm production was not associated with the radiographic profusion categories of simple silicosis or with the occurrence of large opacities.
Some degree of dyspnea was experienced by 125 (87 %) of the patients. In patients with simple silicosis a slight inclination towards a higher degree of breathlessness was observed with advancing radiographic profusion (table 3), which, however, was not significant. All nine patients with large opacities experienced dyspnea, three of grade 2, four of grade 3, and one each of grades 4 and 5.
Lung function in relation to the radiographic findings of the patients with silicosis Table 4 shows the mean values of the lung fun ction tests (reliable spirometric measurements obtained for 139 patients and reliable OLeo measurements for 134 patients) in relation to the radiographic findings of the patients with silicosis. The mean value of the OLeo was lower for the patients in radiographic profusion category 3 than for tho se in category 2 (p < 0.01) or category 1 (p < 0.01), whereas the patient s in categories 1 and 2 did not differ at a statistically significant level. The OLeo was also lower in the patients with than in tho se without large opacities        b Delta % indi cates the positive (+) or negative (-) deviat ion from the predicted FEV % . dieted) was found in 7 % (table 5). Twenty-two (37 %) of the patients with a decreased DL co had no restrictive pattern in their spirogram. In nonsmoking patients with simple silicosis the ve or In patients with large opacities (N = 9) the ve, FEV 1.0' and DL co were less than 80 % of the predicted values in seven, five, and nine patients, respectively. The ve was less than 65 % of the predicted value in two patients. The FEV 1.0 was also less than 65 % of the predicted value in four patients, and the DL co was lowered to a corresponding degree in seven patients. The ventilatory function was obstruc-

Symptoms and lung function in the silicosis patients and their matched referents
The prevalence of persistent phlegm production in the 28 silicosis patients and their referents (43 vs 32 0,10) did not differ at a statistically significant level. The 28 patients with silicosis experienced dyspnea more often than the referents (93 vs 21 %, chi square = 18.2, p < 0.0001).
The silicosis patients had statistically significantly lower mean values for ve, FEVI.O' and DL co (reliable measurements for 26 pairs) than their similarly exposed referents (table 6). The ve, FEVI.o, and DL co were less than 80 % of the predicted values in 9 (35 %),3 (12 %) , and 11 (42 %) of the 26 patients, respectively. Such an impairment was found for ve in three (12 0/0), for FEVl,o in one (4 %), and for DL co in two (8 %) referents, all of whom were smokers or exsmokers.

Discussion
It has been suggested that simple silicosis does not usually produce significant impairment in lung function (4,8,12,14). The present patients with simple silicosis showed, however, a rather high prevalence (46 %) of a restrictive pattern (Ve < 80 % of predicted) in their spirogram. The ve was below 65 % of the predicted value in 15 % of these patients.
A high proportion (47 %) of all the patients with simple silicosis had an impaired DL co « 80 % of predicted), in contrast to some earlier observations (5,13) indicating that the DL co is not impaired in cases of simple nodular silicosis in general. Nine of the 12 patients in category 3 had an impaired DL co, the impairment being below 65 % of the predicted in as many as six patients. This finding suggests that DL co is a rather sensitive lung function parameter for patients with advanced simple silicosis (category 3). Moreover, the mean DL co was lower in patients of category 3 than in those of categories 1 and 2 (65 vs 85 and 82 % of the predicted). The prevalence s of lung function impairment among the nonsmoking patients with simple silicosis rvc 20 %, FEVl,o 20 %, DL co 29 %) were lower than those of patients who were smokers or exsmokers rvc 52 %, FEV l,o 39 %, DL co 51 %). This finding indicates that smoking modified the ventilatory function and diffusion capacity of the patients with simple silicosis. Although found in a relatively small number of nonsmoking patients with simple silicosis, the results further indicate that many nonsmoking patients with simple silicosis have an impaired lung function .
An impairment of spirometric lung function or diffusion capacity or a combination of the two was found in all the patients with large opacities. The result corroborates earlier observations and the suggestions of reviewers (4,5,8,12,13,14).
As noted earlier (9,10), the prevalence of dyspnea on exertion was high among the silicosis patients. Unexpectedly, dyspnea was not associated with the profusion of small opacities, although a trend towards a higher degree of dyspnea was found with more advanced cases of simple silicosis. However, all nine patients with large opacities experienced dyspnea, even grades 4 and 5 (one case each) .
The mean values of ve, FEVI.O' and DL co were lower in the silicosis patients than in their exposurematched referents with similar smoking habits. This finding suggests that silicosis rather than silica dust exposure per se affects pulmonary function. In this respect the results contradict those of Irwig & Rocks (3), who did not find any difference in FVe and FEV1.0 among silicotic gold miners and their referents matched for age, dust exposure, and smoking. Moreover the present patients with silicosis experienced dyspnea more often than their referents. The prevalence of persistent phlegm production was not higher in the patients than in the referents, however, and therefore the phlegm production may not have been caused by the silicotic process, a possibility which conforms with the observation of Irwig & Rocks (3).
Most of the referents (25 out of 28) were currently exposed to silica dust, in contrast to the cases with silicosis (5 out of 28), most of whom had been exposed years ago. Thus one may have anticipated an even higher prevalence of persistent phlegm production in the referents than in the cases with silicosis (32 versus 43 %) had the persistent phlegm production been caused by the silica dust exposure rather than by the silicosis. This result, as well as the pulmonary function findings when the patients and their referents were compared, may, however, have been influenced by the study design. The exposure-matched referents may represent a reference group with a resistant constitution, whereas the cases with silicosis are likely to represent susceptible individuals. The referents may represent a "survival population" with respect to silicosis. Thus one may assume that the referents are more resistant to the fibrogenic effect of silica. Following this line of thought, one may also assume that the referents may be more resistent to other harmful inhalants, including smoking. However, this study design was originally adopted in order that the hereditary characteristics influencing the propensity to contract silicosis could be studied (6).
In conclusion the silicosis patients with simple or complicated (ie, in this case large radiographic opacities) silicosis showed a high prevalence of dyspnea and restricti ve impairment of lung function, as well as a high prevalence of impaired DL co. The DL co proved to be a rather sensitive lung function parameter in cases of advanced simple silicosis. Although smoking clearly modified the effect of silicosis on lung function, one-third of the nonsmoking patients with simple silicosis had an impaired DL co.