Short Lung cancer mortality update and prevalence of smoking among copper miners and smelters

Objectives The aim of this investigation was to study the cancer mortality of Chinese copper miners and smelters further, with particular reference to that fiorn lung cancer, and smoking prevalence. Methods From an earlier follow-up (1970-1985) of the mortality of the two cohorts, all new death cases registered since 1985 were recorded, and the mortality analysis was extended through 1992. A questionnaire survey of smoking habits was carried out in three samples, randomly chosen from the copper miners (N = 1125), smelters (N = 603), and local residents (N = 1517) of Tongling city. Results Lung cancer was significantly increased among the copper miners [standardized mortality ratio (SMR) 152,95% confidence interval (95% CI) 123-1871, but not among the copper smelters (SMR 102,95% CI 53-178). Smoking was more prevalent among copper miners than among local male residents (71.7 versus 64.3%, P < 0.001), whereas among the smelters it was significantly less prevalent (57.4 versus 64.3%, P < 0.005). Similar patterns were found for the average number of cigarettes smoked daily among the miners (21.6 k 7.2), smelters (15 + 7.1), and local Inale residents (19.2 + 7.3). COflcl~SiOflS In addition to occupational exposures, cigarette smoking may partly play a role in influencing mortality from lung cancer among Chinese copper miners and smelters.

(China) between 1970 to 1985. We have previously reported that, compared with the general population of Tongling city, there was an increased risk of lung cancer for the copper miners [standardized mortality ratio (SMR) 147, P < 0.051, but not for the smelters (SMR = 90, P > 0.051, although the overall cancer mortality of the latter group was high (SMR 148, P < 0.05) (1,2). In this paper we report an extension of the mortality analysis up to 1992, together with the results of a survey of the smoking habits of the cohorts and of the general population.

Subjects and methods
Two cohorts of subjects were recruited. The methods have been described in detail elsewhere (1, 2). In brief, the two cohorts consisted of all male workers in copper mines and a smelter factory in Tongling, China. These men had worked for at least one year from 1 January 1969 to 30 June 1985. The occupational records were obtained from the departments of personllel or labor and salary. Information on each subject was retrieved from the records, including name, date of birth, dates of starting and leaving employment, descriptions and dates of all jobs undertaken, and vital status. The death certificates were copied from the local hospitals. Their mortality between 1970 and 1985 was compared with that of Tongling city residents. From 1985, the data bases of two cohorts have been stored in a computer, and all new deaths in the two cohorts have been reviewed prospectively and put into data bases. Standardized mortality ratios were calculated. The statistical significance of the standardized mortality ratio was tested, and the 95% confidence intervals (95% CI) were obtained on the assumption of a Poisson distribution (3).
In 1990-1991, a questionnaire survey of smoking habits was sent to three representative samples, randomly chosen from copper miners, smelters, and local residents of Tongling city. The numbers of subjects were 1125, 603, and 1517, respectively. These subjects were selected by cluster sampling methods (4). The selection was based on a list of the subjects' work units within cohorts. Using such cluster sampling, we could collect the data easily. The subjects were interviewed with questionnaires including name, date of birth, and duration and amounts of cigarettes smoked. The difference in the smoking rates between the groups were examined with a chi-square test (5). We did not include the deceased subjects to avoid some potential recalling bias. The proportion of deaths was 17% for the copper miners and 8% for the copper smelters. We visited the subjects at their homes, if necessary, to obtain a 100% response rate for the questionnaires.

Results
There was an excess of mortality for all sites of cancer among the copper ininers and copper smelters (table 1). The risk of lung cancer among the copper smelters was not significant (SMR 102,95% CI 53-178), while there was a significant increase for liver cancer (SMR 178, 95% CI 105-281). The standardized mortality ratio for lung cancer among the copper ininers was significantly elevated to 152, and the miners also had an increased

Discussion
The Tongling copper mine is one of the largest and oldest copper mines in China. It was expanded to largescale production after 1950. In 1953, the Tongling smelter factory was built to accommodate the copper mineral ores. During the 1950s, both of these facilities employed a larger number of workers, administered by the Tongling color metal company. These workers had almost the same characteristics, such as birth place, education, economic status, age, and year at the time of starting work in the mines and factory. They were the majority of subjects in the two cohorts, respectively. Therefore a comparison of the standardized mortality ratios of the  two groups seemed suitable. In the underground of the mines, the level of radioactivity was 0.55 . lo-" curie. 1-I, which was under the occupational exposure limit of China, and the dust density was approximately 10 mg . m-' on the average. Earlier estimates of the arsenic level in the work environments of the smelter factory showed no significant difference from the local living environment; the Tongling copper mineral ores contained almost no arsenic. But after the 1970s, the ores from other copper mines were smelted in this factory, and the arsenic level showed an increase of two to four times in this work environment over a control area. By the end of the follow-up, it was found that 368 miners and 141 smelters could not be traced and the rates lost to follow-up were 5% for the miners and 7% for the smelters. In the investigation of workers' smoking habits, however, we were able to interview all these subjects, including the retired ones who were visited in their homes, to avoid any loss. The questionnaire study included the living subjects; it used no data on the smoking habits of the subjects who died before the investigation in order to decrease some recalling bias. Nevertheless, the randomized sampling methods were the same for all three groups, and the smoking habits of these subjects should still be comparable. The results reflect the differences of the smoking habits between the copper miners and smelters.
In earlier results from 1970 to 1985, the mortality from lung cancer among copper smelters was not high (5 observed versus 5.56 expected), probably because the cohort was younger. This cohort was followed through 1991, but no significant change in the mortality pattern of lung cancer was observed among these workers. We found that there was a higher risk of lung cancer among the copper miners (89 observed versus 58.55 expected). The employees in the Tongling copper mines and smelting factory could smoke freely. When the copper smelters were working, however, smoking was restricted due to the evil-smelling fumes from copper smelting. In addition, the copper miners tended to smoke more frequently during inactive periods underground, the hazards of smoking could have been increased because of poor ventilation in the underground workplace.
Cigarette smoking is a risk factor for lung cancer, chronic obstructive pulmonary disease, coronary heart disease, and stroke (6). It is commonly believed that men in Asian countries tend to smoke heavily, and the prevalence of smoking could reach 68.9% in China (7). Numerous studies have shown that cancer caused by occupational exposure could be colnpounded by smoking (8)(9)(10)(11)(12). A recent cohort study of occupational cancer in China indicated that smoking can synergistically increase the risk of occupational lung cancer (13). Many epidemiologic studies of copper smelters have shown an excess of lung cancer (14)(15)(16)(17)(18). The Swedish results revealed an increased risk of lung cancer in the lowest arsenic exposure categories. A higher proportion of cigarette smoking among these smelters may have contributed to these observations (1 [8][9][10][11][12][13][14][15][16][17][18][19][20]. Arsenic exposure in one copper-smelting factory was considered a factor promoting cancer (17). Our study showed a lower prevalence of smoking among smelters than that of the comparison group. Perhaps this lower prevalence could explain the fact that there was no excess of lung cancer mortality among our copper smelters. In other words, it could have led to negative confounding (21). Some studies have indicated that radioactivity can be a risk factor for lung cancer among metal miners (8,10,22,23). In the Tongling copper mines, the level of radioactivity was below the occupational exposure limit (2). These miners were exposed mainly to silica dust. Silica is suspected of being a carcinogen (24). In addition to the occupational exposures (25), we believe that cigarette smoking could have played a role in the excess of lung cancer due to the high prevalence of smoking found among the copper miners, although this effect of smoking on the lung cancer of the miners may be partial because it did not entirely account for the standardized mortality ratio being more than 150 for lung cancer (21). Therefore, a nested case-referent study needs to be carried out to estimate the exact effects the smoking had on lung cancer among the copper miners. The results of the smoking habits of the copper miners and smelters in this investigation could provide clues to their different mortality from lung cancer. A campaign promoting smoking cessation, especially in workshops, should be advocated for workers in the Tongling copper mines.