A cross-sectional study was carried out at a coal mine in Mbeya, Tanzania in 2003 and 2004. Of the 556 workers in this mine, 220 workers were excluded. The excluded workers were managers, assistant managers and heads of section due to their high socio-economic status. In addition surface workers in carpentry, masonry, garage, foundry, welding, machine workshop and surveying were excluded due to other types of dust exposure. Office workers and temporary workers were also excluded. Thus, 336 workers were invited to participate; 318 participated (303 men and 15 women), giving a response rate of 94.6%. The women were excluded before the statistical analysis due to their low number, as well as two workers with bronchial asthma and two with tuberculosis. The remaining 250 workers from the production part of the mine constituted the final study population. These were high-exposure workers from the development team (n = 47) and lower-exposure workers from the mine team (n = 78) and from the other production teams (n = 125). The tasks for the above teams are described in our previous publication (19).
The coal mine workers were interviewed using a respiratory health questionnaire. The questionnaire had three parts, including personal and work characteristics, smoking habits and respiratory health symptoms. The questionnaire was prepared in English and was translated into Swahili, the national language of Tanzania, it was used in the previous study. The questionnaire was pre-tested among 30 selected coal mine workers and discussed for clarity before the study started. The questions on personal and work characteristics included sex, age, education level, employment history, years worked in the mine and years in dusty work elsewhere. The questionnaire was administered between 0800 and 1600.
Acute symptoms were assessed using a modified optimal symptom score questionnaire  and scored on a five-point Likert scale as never (1), mild (2), moderate (3), severe (4) or very severe (5). Workers were asked whether they had the following symptoms: dry cough, shortness of breath, wheezing, stuffy nose, runny nose and sneezing during or after the previous shift. Before statistical analysis, the responses were dichotomized to no (never) and yes (mild, moderate, severe or very severe).
A modified version of the British Medical Research Council questionnaire on respiratory symptoms  included questions on whether respondents usually had symptoms of cough, breathlessness and wheezing. The subjects were also asked whether they had bronchial asthma and/or other chronic illnesses such as tuberculosis and bronchitis (yes/no). Further, the workers were asked whether they had had injuries or surgery affecting the chest and whether they had had heart problems, pneumonia, pleurisy, pulmonary tuberculosis, bronchial asthma or any other chest problems in the past 3 years (yes/no). Those with any of these problems were excluded from the analysis.
Current smokers were defined as those who were smoking at the time of the study or those who had smoked more than one cigarette per day and stopped less than 1 year prior to the study. Ex-smokers were those who had smoked previously and stopped more than 1 year previously. The year they stopped smoking and the numbers of cigarettes smoked per day were also recorded. Never-smokers were defined as individuals who had never smoked.
Assessment of exposure
As part of our previous exposure assessment , carried out concomitantly with the presently reported questionnaire studies on respiratory symptoms, personal dust was sampled during the day shift, which normally lasted about 5–10 hours. Five full-shift samples were taken on each monitoring day. Personal respirable dust was sampled using a SKC Sidekick pump (model 224–50) with a flow rate of 2.2 l/min. A rotameter was used to adjust the flow. The respirable dust samples were collected on 37-mm cellulose acetate filters (pore size 0.8 μm) placed in a 37-mm conductive plastic cyclone. The cassette was assembled and labeled at X-lab in Bergen, Norway. The cyclone was clipped to the worker's collar, allowing it to hang freely and collect dust in the breathing zone.
The respirable dust samples were quantified by gravimetric analysis using a Mettler AT 261 delta range with a limit of detection of 0.01 mg/m3. Respirable dust samples were analysed for quartz by X-ray diffraction on a silver membrane filter using NIOSH method 7500 at SGAB Analytica Laboratory, Luleå, Sweden. The limit of detection was 0.005 mg/m3 .
Cumulative dust exposure
The individual cumulative exposure to respirable dust or quartz (mg·year/m3) for the workers was estimated as the sum of the product of the estimated worker-specific mean exposure in the respective job teams and number of years the worker had spent in these job teams .
The Statistical Package for the Social Sciences (SPSS) version 12 was used for the data analysis. P ≤ 0.05 was chosen as the criterion for statistical significance. The independent t-test was used to compare continuous variables between the development, the mine and the other production workers. The chi-square test was used to compare proportions in categorical variables. Logistic regression analysis was used for groups where the number of workers with symptoms are about 15  to determine odds ratio (OR) for groups with chronic respiratory symptoms based on quartiles and the highest deciles of cumulative exposure using the lowest quartile as the reference group, while adjusting for ever-smoking and age.
Summary variables for chronic symptoms and for acute symptoms were created to evaluate the correlation between chronic and acute respiratory symptoms. For chronic respiratory symptoms this was created by summarizing the score of each symptom; to have cough first thing in the morning, cough during the day and night, cough with sputum first thing in the morning, cough with sputum during the day and night, shortness of breath when hurrying on level ground and shortness of breath walking with people of your own age on level ground. This sum score ranged from 0 to 6. Summarizing the scores for dry cough, shortness of breath, wheezing, stuffy nose, runny nose and sneezing, created the summary variable of the acute respiratory symptoms with score (0–5). Pearson correlation coefficients were calculated for estimating the correlation between acute and chronic symptoms.
Ethical approval and informed consent
Ethical approval was obtained from the Western Norway Regional Committee for Medical Research Ethics and the National Institute for Medical Research of Tanzania. The research permit was obtained from the Tanzania Commission for Science and Technology (COSTECH). There was institutional consent, since the administration of the Kiwira Coal Mine was informed of the project and allowed the study to proceed. Each person was informed about the aims of the study and the methods before being requested to consent to participate in the study voluntarily.