Occupation and testicular cancer.

smokers of low tar cigarettes have an estimated "intake" of tar some 25% lower than that of smokers of higher yielding brands. They emphasise, as has been previously reported,' that when making interproduct comparisons and extrapolating from one smoke marker to another, it is important to take into account differing ratios ofthese markers (as measured under standard conditions) for the products being compared. In a very different study2 of demographically representative samples of regular middle tar and regular low tar smokers, I found that estimates of mouth tar intake (derived from 24 hour butt nicotine analyses) indicated a 32% lower intake for the low tar smokers. The results also showed a similar reduction in tar intake (30%) for the middle tar smokers when they were asked to smoke a low tar cigarette. The indirect methodology employed in that report has now been validated through a detailed experimental study (RGR, in preparation). Russell's study examined the tar intake of heavy smokers (around 30 cigarettes per day) and he was cautious in extrapolating his findings to lighter smokers. However, the low tar smokers in my study2 consumed on average 17-5 cigarettes a day, which indicates that consumption is not a critical factor in determining smoking patterns. Russell's results are thus supported in studies of (a) widely differing design and methodology, (b) both representative and non-representative populations of smokers, and (c) both habitual low tar versus habitual middle tar smoking and acute switching between middle tar and low tar smoking. The fact that these studies indicate comparable results supports the validity of the overall conclusion that low tar means less tar.

They emphasise, as has been previously reported,' that when making interproduct comparisons and extrapolating from one smoke marker to another, it is important to take into account differing ratios ofthese markers (as measured under standard conditions) for the products being compared.
In a very different study2 of demographically representative samples of regular middle tar and regular low tar smokers, I found that estimates of mouth tar intake (derived from 24 hour butt nicotine analyses) indicated a 32% lower intake for the low tar smokers. The results also showed a similar reduction in tar intake (30%) for the middle tar smokers when they were asked to smoke a low tar cigarette. The indirect methodology employed in that report has now been validated through a detailed experimental study (RGR, in preparation).
Russell's study examined the tar intake of heavy smokers (around 30 cigarettes per day) and he was cautious in extrapolating his findings to lighter smokers. However, the low tar smokers in my study2 consumed on average 17-5 cigarettes a day, which indicates that consumption is not a critical factor in determining smoking patterns.
Russell's results are thus supported in studies of (a) widely differing design and methodology, (b) both representative and non-representative populations of smokers, and (c) both habitual low tar versus habitual middle tar smoking and acute switching between middle tar and low tar smoking.
The fact that these studies indicate comparable results supports the validity of the overall conclusion that low tar means less tar. ROGER  1986: 40, 26-9) were trying to determine whether an observed association between occupation and testicular cancer could highlight aspects of modern 279 living that may contribute to an increased incidence of the disease. Deriving their data from mortality statistics (ICD:9 186), they were unable to differentiate between the two main histological categories of testicular cancer: seminoma and teratoma. The very different pathologies of seminoma and teratoma and the differences in their epidemiological characteristics probably suggest different aetiological factors. (Seminomas occur more frequently than teratomas, their incidence peaks earlier, and a larger percentage occur in social clases I and III).
Although the authors did mention the possible confounding influence of survival time, this was in relation to testicular cancer en masse, and they did not indicate what type of cancer patient was surviving. The treatment of seminomas has remained fairly constant over their study period (1971-80), orchidectomy and/or orchidectomy with radiotherapy constituting standard procedure. It is likely therefore that the type of seminoma patient dying has also remained constant. The same is not true, however, for teratoma patients. In 1975, marked progress in cytotoxic chemotherapy, particularly in the use of cisplatin in the treatment of germ cell tumours, increased the percentage of those surviving from malignant teratoma (and seminoma with extralymphatic metastatic disease) from 8% to 90%.2 Thus, in the study period the composition of patients will have altered appreciably.
A scrutiny of the death certificates to identify the occupations separately for patients with teratomas and seminomas would be an important next step.

Testis cancer
McDowall and Balarajan' describe an increasing incidence oftestis cancer in young men in England and Wales; they later mention an increased mortality. Their premise for concluding an increased incidence is based on the assumption of rising mortality rates as previously reported.2 This parallel between increasing mortality and incidence may have been true when the original paper by Davies was published, but the fact is that mortality from testis cancer in England and Wales is now decreasing. In a paper in this Journal, we