We analysed the relation between coffee consumption and hepatocellular carcinoma in two case-control studies conducted between 1984 and 1998 in Italy and Greece, including 834 cases and 1912 controls. Compared to non coffee drinkers, the multivariate odds ratio was 0.7 for drinkers of three or more cups per day.

The relationship between physical activity and the risk of ovarian cancer was analyzed using data from a case-control study conducted between 1992 and 1999 in Italy. Cases were 1,031 women with incident, histologically confirmed, invasive epithelial ovarian cancer and controls were 2,411 women admitted to hospital for acute non-neoplastic, non-hormonal conditions. Compared to women with the lowest level of occupational physical activity, the ORs of ovarian cancer obtained adjusting for center, year of interview and age for women with the highest level of physical activity were 0.70, 0.52 and 0.64 (all statistically significant) respectively, at ages 15-19, 30-39 and 50-59 years, with significant trends in risk for the 2 youngest age groups. The corresponding ORs became 0.89, 0.67 and 0.76 after further allowance for several co-variates of ovarian cancer, including education, which was positively associated with cancer risk. No significant association was found with leisure-time physical activity though the risk was below unity in women with the highest level of activity. The multivariate OR was 0.44 for women with the highest level of combined occupational plus leisure-time physical activity. The inverse relationship between occupational physical activity and ovarian cancer risk was not heterogeneous across strata of selected co-variates.

EDITOR—Marmot and Bobak analysed the increased inequalities in health in eastern Europe.1 Cervical cancer is an avoidable cause of death and a relevant indicator of women's health. National death certification data do not allow analysis of mortality from cervical cancer in Europe since 20-65% of deaths from uterine cancers are certified reliably as uterus, unspecified.2 Most deaths from uterine cancer in women aged under 45 arise from the cervix. We analysed age standardised death certification rates from uterine cancer in women aged …

After the peak rate of cancer mortality reached in 1988 in the European Union, steady declines were observed: 9.1% for both sexes combined over the period 1988-1997 (from 147.0 to 133.6/100,000, world standard), corresponding to the avoidance of about 80,000 deaths in 1997 (approximately 39,000 below age 65 and 41,000 above). In 1997, the total number of cancer deaths also declined, for the first time. The major determining cancers for these favorable trends were stomach (-30%), lung (-10%), intestines (-15%), breast (-10%), uterus (mainly cervix; -22%), leukemias (-10%) and, after 1995, prostate (-3%).

The relation between diabetes and prostate cancer risk was investigated in a combined analysis of two hospital-based case-control studies conducted in Italy and Greece, between 1985 and 1997. Cases were 608 men with incident prostate cancer, and controls were 1008 men admitted to hospital for acute non-neoplastic diseases. No material association between diabetes and prostate cancer was observed, with a multivariate odds ratio (OR) of 1.07 (95% confidence interval (CI) 0.68-1.66). Compared with men without diabetes, there was a non-significant increased risk of prostate cancer in those diagnosed with diabetes within the last 5 years (OR 2.04), while the ORs were 0.96 and 0.78 respectively for a diagnosis of diabetes 5-9 years and > or = 10 years ago.

The influence of food groups on oesophageal cancer risk was analysed using data from a case-control study conducted between 1992 and 1999 in the Swiss Canton of Vaud on 101 incident, histologically confirmed cases (92 squamous cell, 9 adenocarcinomas) and 327 controls admitted to hospital for acute, non-neoplastic conditions. Multivariate odds ratios (OR) were computed after allowance for age, sex, tobacco, alcohol and non-alcohol energy. Significant increased risks emerged for red meat (OR = 1.7 for an increase of one serving per day), pork and processed meat (OR = 1.6), and eggs (OR = 1.5), whereas inverse associations were observed for milk (OR = 0.7), raw and cooked vegetables (OR = 0.5), citrus and other fruits (OR = 0.5), as well as for a more varied diet (OR = 0.5). Most associations were apparently stronger in heavy alcohol drinkers, suggesting an interaction between poor diet and alcohol drinking in oesophageal carcinogenesis.

Although the use of hormone replacement therapy for menopause based on estrogens alone increases endometrial cancer risk, several studies indicate that combined therapy with estrogens and progestins is not related to an appreciable excess of endometrial cancer risk, if progestins are given for more than 10 or 14 days in each cycle. Combined therapy, however, may be associated to greater excess in breast cancer incidence as compared to estrogens alone.

Hodgkin's disease mortality rates steadily declined by about 75% between the late 1960's and the late 1990's in the current European Union countries and the USA, and Japan. Eastern European countries, however, showed only an approximately 40% decline between the late 1960's and the early 1990's, and no further fall thereafter.

The reasons for the worldwide decline in stomach cancer incidence and mortality rates are not fully understood, but dietary changes are clearly implicated. While the possible mechanisms of gastric carcinogenesis and the impact of Helicobacter pylori eradication remain open to debate, at least two practical recommendations - to increase fruit and vegetable intake, and to reduce consumption of salt - are already supported by epidemiological evidence. These dietary recommendations may also be beneficial in the prevention of other cancers and chronic diseases. Promising evidence of a favourable effect of certain vitamins, such as vitamin C and E and beta-carotene, and minerals, such as selenium, justifies additional investigation.

Post-menopausal women who have never used hormone replacement therapy have a higher risk of colon, but not rectal, cancer than do premenopausal women of the same age, socio-cultural class and dietary habits. Such risk increase seems to last about 10 years and to be restricted to lean women, a group who have lower levels of oestradiol after ovarian function ceases after menopause.

In a cohort mortality study, a few basic rules must be respected for risk estimation including: (1) the definition of the criteria of entry in the cohort; (2) the time period of observation; (3) the outcome, which must be the same for observed and expected events [Breslow and Day, 1987; MacMahon and Trichopoulos, 1996]. In their letter to the Editor referring to our article "Mortality Among Workers in the Geothermal Power Plants at Larderello, Italy" [Pira et al., 1999], Merler et al. appear to ignore these basic principles, since they extend the observation period from 1990 to 1998 for observed but not for expected events and do not distinguish between incidence and mortality. Consequently, their relative risk estimates are simply not interpretable. Anyway, an update of the follow-up to the end of 2000 has already been planned.