The distribution of cases and controls according to place of birth, education, heating fuel, work environment, tobacco use, and alcohol consumption is shown in and .
The risk for oesophageal, lung, oral and laryngeal cancers in males in relation to selected risk factors
The risk for oesophageal, lung and oral cancers in females in relation to selected risk factors
Women who were born in the Eastern Cape Province had a significantly increased risk [Odds Ratio (OR)=6.1] of being diagnosed with oesophageal cancer compared to those born elsewhere in South Africa or outside of the country. When duration of residence in the Eastern Cape was looked at () length of residence had a large effect for women (OR=3.6 for 1–34 years; OR=14.7 for 35 or more years) but was only marginally significant for men who had lived there for 35 or more years (OR=3.1).
The risk for oesophageal cancer in relation to duration of residence in Eastern Cape province, South Africa
Compared with eight or more years of education, lower levels were associated with elevated odds ratio estimates for oesophageal cancer of 2.0 for men with no education, 1.6 for men with only primary education, and 2.2 for women with no education. Oral cancers were about four times more frequent in women with no or only primary education (OR=4.2 and 3.9), compared to those with secondary or higher education; this difference was not found for these cancers in men. However males with no education were more likely to be diagnosed with laryngeal cancer (OR=3.3).
Using ages 18–44 as the base line, in women there was a clear trend for increasing risk of oesophageal cancer with increasing age; the trend was less clear for men where 55–64 year olds have the greatest estimated risk. With respect to lung cancer, women 65–74 had approximately twice (OR=11.4) the estimated risk compared to those aged 45–64. This was in contrast to men whose risk was three to four times base line at all ages 45 and over. Age was not a risk factor for oral cancer in women and possibly not for laryngeal cancer in men (only showing a significant increase in men aged 45–54); however it was significantly associated with oral cancers in men 45–64 years old.
At the time of their interview 73% of controls reported using electricity for both cooking and heating, with only 18% using wood, charcoal, coal, or anthracite. In the past the situation was quite different with 82% reporting the use of the latter fuels, and only 15% using electricity ‘20 years ago’.
The reported use of wood, charcoal, coal or anthracite for heating ‘20 years ago’ was not associated with any significant increase in the cancers studied. However, although the numbers were small, men, but not women, who reported using liquid paraffin (kerosene) for heating were at greater risk for lung cancer (OR=5.5) and women, but not men, using this fuel were at increased risk for oesophageal cancer (OR=3.5).
Using broad groupings for industrial/workplace classifications, males working in areas with ‘potentially noxious’ exposures had increased risks: they were 2.9 times more likely to be diagnosed with lung cancer than their counterparts in administrative, clerical and sales businesses. There were too few women (n=38) working in ‘potentially noxious’ environments to draw inferences.
Tobacco smoking, past or current, was the major risk factor for all of the cancers included in this study. For cases the mean duration of smoking was 22.1 (s.d.±19.0) years for males and 6.4 (s.d.±14.3) years for females, compared to 3.5 (s.d.±9.8) and 0.8 (s.d.±5.0) years respectively for controls.
Ex-smokers and ‘light’ current smokers had about a three-fold increased risk for oesophageal cancer, being slightly higher for men than for women. For ‘heavy’ smokers the risk doubled to an OR of about six in both sexes.
Lung cancer was six to 13 times more likely to be diagnosed in ex- or ‘light’ smokers, with the risk being slightly higher in women than in men. In the case of ‘heavy’ smokers the odds ratios were 23.9 in men and 50.9 in women, both with wide confidence limits.
The increased risk of oral cancers in women who smoked was similar to that for oesophageal cancers; however, for men it was somewhat increased to an OR of 12.5 in ‘heavy’ smokers.
Almost all of the men (49/51) with laryngeal cancer had smoked at some time, with increased risks ranging from an OR of 10.2 for ‘light’ smokers to an OR of 23.6 for ‘heavy’ smokers.
Snuff use was more common among women than smoking, with 22% of female control patients and 4% of males reporting this habit. ‘Ever’ use of snuff as compared to ‘never’ use did not appear to be associated with an increased risk of oesophageal, lung, or oral cancer.
Frequency of alcohol consumption, on its own, was not a major contributor to the cancers studied. The only significant increased risk was for oesophageal cancer in women classified as ‘weekly drinkers’ (OR=2.2) and women (OR=1.7) and men (OR=1.8) classified as ‘frequent drinkers’. In addition, when alcohol consumption was combined with smoking the risk of the association with oesophageal cancer was significantly increased (OR=4.4 men and women combined; see ).
The risk for oesophageal cancer in males and females in relation to the combined effect of tobacco and alcohol consumption