The relation between selected aspects of diet and the risk of oral and pharyngeal cancer was considered in a few cohort studies and approximately 30 case-control studies. These studies reported consistent inverse associations with fruit and vegetable consumption. beta-carotene, vitamin C and selected flavonoids were also inversely related to risk, although it remains difficult to disentangle their potential effect from that of fruit and vegetables. Whole-grain cereals, but not refined grain ones, were also favorably related to the risk of oral and pharyngeal cancer. The results were not consistent for other foods, including meat, fish, eggs, milk and dairy products, but it is now possible to exclude a strong relation with oral and pharyngeal cancer risk. Data are also reassuring for coffee and tea, while hot maté drinking has been related to increased risk in studies from Argentina and Brazil. There is no consistent association with total fat intake, but monounsaturated fats (and olive oil) have resulted inversely related to risk. In developed countries, selected aspects of diet may account for 20% to 25% of oral and pharyngeal cancers. This proportion is likely greater in selected developing countries.

Mortality from gallbladder cancer has been traditionally high in Eastern Europe, and lower in northern countries. Trends in 18 European countries, including the European Union (EU) and selected Eastern European countries, have been updated using official death certification data abstracted from the WHO database over the period 1980-1999. In the EU, age-standardized rates declined by about 30% between the late 1980s and 1999 to reach 1.8/100 000 for women, and by about 10% to reach 1.4/100 000 for men. In the Czech Republic and Hungary, rates for women were over 6/100 000 until the early 1990s, and declined by about 25% thereafter. For males, gallbladder cancer mortality showed no consistent trend, with rates over 3/100 000. Thus, a high mortality area from gallbladder cancer is still evident for both sexes in Central and Eastern Europe. The trends in mortality from gallbladder cancer are probably influenced by changes in risk factor exposure, such as diet, nutrition or tobacco, but essentially reflect more widespread and earlier adoption of cholecystectomy in the EU, since gallstones are the major risk factor for gallbladder cancer. The data also indicate the scope for further improvement of the management of gallbladder disease in Eastern Europe.

The relationship between hormonal therapy for menopause (hormone replacement therapy, HRT) and the risk of epithelial ovarian cancer was evaluated in a collaborative re-analysis of 4 European case-control studies, 2 conducted in Greece and 1 each in Italy and the United Kingdom, including a total of 1,470 ovarian cancer cases and 3,271 hospital controls. Odds ratios (ORs) for HRT use were derived after allowance for study centre, age, socio-economic level, parity, menopausal status, type of menopause, age at menopause and oral contraceptive use. Overall, 109 (8.0%) ovarian cancer cases and 146 (4.7%) controls had ever used HRT, corresponding to an adjusted OR of 1.71 (95% confidence interval 1.30-2.25). The point estimates of the OR were 1.77 in the first Greek study, 1.40 in the second Greek study, 1.66 in the Italian study and 1.68 in the British study. Adjustment for possible confounders, including menopausal status, type of menopause, age at menopause and oral contraceptive use, slightly increased the OR. Limiting the analysis to women with information on relevant aspects of HRT use revealed a weak positive association with duration and some evidence that the excess relative risk for ovarian cancer declined with time since last use. These findings are compatible with a promoting effect of HRT in ovarian carcinogenesis. It is also possible, however, that the positive association reflects chance or selective administration of HRT to high-risk individuals, since until recently in Europe HRT was prescribed mainly for alleviation of peri-menopausal symptoms.

Death certification data from pleural cancer in eight European countries providing data to the World Health Organization database over the period 1970–1994 were analysed using a log-linear Poisson model to disentangle the effects of age, birth cohort and period of death. The age effect reached values between 10 and 15/100 000 males at age 80–84 in most countries, except Hungary (6.7), Switzerland (18.0), France (20.6) and the Netherlands (36.5). Cohort effects were steadily and appreciably upwards in all countries up to the generations born in 1940 or 1945, and levelled off for the 1950 cohort, except in Hungary, where persistent rises were observed. Thus, most rises in pleural cancer mortality in Europe were on a cohort of birth basis. Since most pleural cases were asbestos-related mesotheliomas, and since asbestos has an early-stage effect on subsequent mesothelioma risk, exposure early in life is important for determining the subsequent mesothelioma risk of each generation. Consequently, the data indicate that the peak mortality from pleural cancer in most western European countries will be reached in the first decades of the 21st century, i.e. around 2010–2020, when the generations born between 1940 and 1950 will reach the peak age for mesothelioma incidence and mortality. This contrasts with US data, where the peak of pleural cancer incidence has been reached at the end of the 20th century, and reflects a delay in adopting adequate prevention measures since the 1940–1945 generations entered the workforce in the 1960s, when cancer risk from asbestos exposure was already recognized.

A few studies have investigated whether the risk of laryngeal cancer depends on the types of alcoholic beverage consumed, providing conflicting results. We investigated this issue using the data from two case-control studies conducted in Italy between 1986 and 2000. These included 672 cases of laryngeal cancer and 3454 hospital controls, admitted for acute, non-neoplastic conditions, unrelated to smoking and alcohol consumption. Significant trends in risk were found for total alcohol intake, with multivariate odds ratios (ORs) of 1.12 for drinkers of 3-4 drinks/day, 2.43 for 5-7, 3.65 for 8-11, and 4.83 for > 12 drinks/day, as compared to abstainers or light drinkers. Corresponding ORs for wine drinkers were 1.12, 2.45, 3.29 and 5.91. After allowance was made for wine intake, the ORs for beer drinkers were 1.65 for 1-2 drinks/day, and 1.36 for > or = 3 drinks/day, as compared to non-beer drinkers; corresponding values for spirits drinkers were 0.88 and 1.15. This study thus indicates that in the Italian population characterized by frequent wine consumption, wine is the beverage most strongly related to the risk of laryngeal cancer.

A clinical series from the Milan National Cancer Institute,1 as well as the large database of the Danish Breast Cancer Cooperation Group (DBCG)2 indicated that breast cancer localized in the upper lateral quadrant had better prognosis than those located elsewhere in the breast. Nodular status, tumor size, histological grading and patients' age were unable to explain the difference in survival. Because the differences were relatively limited (of the order of 20%), although of potential clinical relevance, independent confirmation from other population-based series would be of importance. We have therefore considered the influence of tumor location on breast cancer survival using data from the Swiss Cancer Registry of Vaud. The catchment population of the Vaud Cancer Registry numbered 602,000 in 1990. Population-based incidence data have been available since 1974. The registry is tumor-based and multiple primary malignancies found in the same person are entered separately. The basic information available for each comprises socio-demographic characteristics of the patient (age, gender), the primary site and histological type of the tumor according to the standard International Classification of Diseases for Oncology, Ninth Revision, (ICD-O. 9)3 and the date of the diagnosis.4 The present series includes 4,562 patients with breast cancer diagnosed between 1989 and 1999. These women's records were actively followed-up through to the end of 1999 to check for emigration or death.5 The product limit (maximum likelihood) method was used to obtain survival rates. Moreover, multivariate hazard ratios, as estimators of relative risk (RR), (and corresponding 95% confidence intervals [CI]) of death are given. Included in the proportional hazard models were terms for nodal status, tumor size (<2, 2 to ≤5, >5 cm), histological grading (I, II + III, other or unknown), age at diagnosis (in years) and year of diagnosis. Table I gives the distribution of 4,562 incident breast cancer cases according to location and selected tumors' characteristics. Tumors located in the lateral quadrant were more frequently nodal positive than medial ones, whereas no difference was observed with reference to tumours' size and histological grading. Five-year crude survival was 78% for neoplasms in the upper lateral quadrant, 79% for those in the lower lateral, 77% in the upper medial and 70% in the lower medial. The difference was statistically significant (p = 0.018). Table II gives the fully adjusted relative risk of dying of breast cancer according to tumor quadrant location. As compared to tumors located in the upper lateral quadrant, the RR was 1.23 (95% CI = 1.07–1.43) for other locations. The RR was 1.20 for lower lateral, 1.02 for upper medial and 1.55 for lower medial. A tendency for lower survival in lower and medial quadrants was also observed across strata of nodular status, although the random variation was greater due to small absolute numbers, The present, population-based dataset therefore confirms that, although crude survival is not consistently related to quadrant location, breast cancer prognosis is more favorable for neoplasms in the upper lateral quadrant than for other locations, when allowance is made for nodal status and other tumor characteristics. This may be related to differences in lymphatic drainage, incomplete removal of tumor tissue in women with tumors distant from the axilla, or to other factors of prognostic importance.2 The confirmed observation of these prognostic differences indicates that tumor location should be included in any multivariate model as an additional factor in breast cancer management and treatment. Yours sincerely, The contribution of the Vaud Cancer Registry's staff is gratefully acknowledged. Fabio Levi, Lalao Randimbison, Van-Cong Te, Carlo La Vecchia

We used data from a multicentre case-control study conducted in Italy between 1991 and 1994 on over 2500 cases of breast cancer and a comparable number of controls, and estimates of breast cancer incidence in Italy to compute individual breast cancer risk for Italian women. The estimated probabilities between age 50 and 80 ranged from approximately 5% (for a woman with no family history and low modifiable risk profile) to about 30% (for a woman with young family history and high modifiable risk) on the basis of various women's baseline characteristics. Expected numbers of breast cancer cases using the present model were compared with those based on the USA Gail model, and with the observed ones in the comparison group of the Italian Tamoxifen Trial. These show a closer agreement between the observed and the expected total numbers of breast cancers than the USA Gail model. Thus, the Gail model can be improved for use in other populations by using estimates of incidence and risk which are more appropriate to the target population.

Cancer has been a growing public health problem throughout the last century. Cancer mortality rates in both sexes combined in the European Union (EU) reached a peak in 1988. Thereafter, they declined by 9.4% in 1998. Likewise, lung cancer rates in the EU for both sexes combined increased by 58% between 1960 and 1988, but declined by 14% in 1998. Over a third of the decline was accounted for by lung cancer alone and approximately half by the combination of tobacco-related neoplasms. About half of the decline in total cancer mortality not attributable to tobacco derived from the steady fall in mortality from gastric cancer. The remaining half, including favourable trends in colorectal, breast, testis and lymphoid neoplasms, can be at least in part attributed to advancements in cancer diagnosis and treatment. The major causes of cancer and hence the most important priorities for research will be reviewed, with a specific focus on European priorities for research.

We considered epidemiological data on overweight, diabetes, insulin, and breast cancer. Overweight is inversely related to premenopausal breast cancer, but there is definite evidence that, as compared with normal weight women, the relative risk (RR) for postmenopausal breast cancer is around 1.5 for overweight women and >2 for obese women, and that the association is stronger in elderly women. Overweight and obesity are strongly related to diabetes. Diabetes is associated with postmenopausal breast cancer, too, with summary RRs from meta-analyses of 1.15-1.20, but not with premenopausal breast cancer (RR, 0.9). There is no consistent evidence that fasting insulin is related to breast cancer risk. Thus, although overweight and obesity are strongly related to postmenopausal breast cancer, diabetes is only moderately related to it. Given the extent of the association, and the likely residual confounding by overweight, inference on causality for the diabetes-breast cancer relation remains open to discussion.

Letters7 October 2003Alcohol, Postmenopausal Hormones, and Breast CancerSilvano Gallus, ScD, Silvia Franceschi, MD, and Carlo La Vecchia, MDSilvano Gallus, ScDFrom Istituto di Ricerche Farmacologiche Mario Negri; 20157 Milan, Italy; International Agency for Research on Cancer; 69372 Lyon Cedex, France; and Istituto di Statistica Medica e Biometria, Universit degli Studi di Milano; 20133 Milan, Italy.Search for more papers by this author, Silvia Franceschi, MDFrom Istituto di Ricerche Farmacologiche Mario Negri; 20157 Milan, Italy; International Agency for Research on Cancer; 69372 Lyon Cedex, France; and Istituto di Statistica Medica e Biometria, Universit degli Studi di Milano; 20133 Milan, Italy.Search for more papers by this author, and Carlo La Vecchia, MDFrom Istituto di Ricerche Farmacologiche Mario Negri; 20157 Milan, Italy; International Agency for Research on Cancer; 69372 Lyon Cedex, France; and Istituto di Statistica Medica e Biometria, Universit degli Studi di Milano; 20133 Milan, Italy.Search for more papers by this authorAuthor, Article, and Disclosure Informationhttps://doi.org/10.7326/0003-4819-139-7-200310070-00022 SectionsAboutFull TextPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinkedInRedditEmail TO THE EDITOR:In the Nurses' Health Study cohort, Chen and colleagues (1) found an approximately 2-fold excess risk for breast cancer in women who currently used postmenopausal hormones and drank alcohol. However, the researchers were unable to consider high alcohol consumption. To address this issue, we analyzed data from 2 Italian case–control studies. The first was conducted between 1983 and 1991 in greater Milan (2), and the second was conducted between 1991 and 1994 in 6 centers in various regions of Italy (3). The studies involved 3573 postmenopausal women (median age, 61 years [range, 31 to 74 years]) with ...References1. Chen WY, Colditz GA, Rosner B, Hankinson SE, Hunter DJ, Manson JE, . Use of postmenopausal hormones, alcohol, and risk for invasive breast cancer. Ann Intern Med. 2002;137:798-804. [PMID: 12435216] LinkGoogle Scholar2. La Vecchia C, Negri E, Parazzini F, Boyle P, Fasoli M, Gentile A, . Alcohol and breast cancer: update from an Italian case–control study. Eur J Cancer Clin Oncol. 1989;25:1711-7. [PMID: 2632254] CrossrefMedlineGoogle Scholar3. Ferraroni M, Decarli A, Franceschi S, La Vecchia C. Alcohol consumption and risk of breast cancer: a multicentre Italian case–control study. Eur J Cancer. 1998;34:1403-9. [PMID: 9849424] CrossrefMedlineGoogle Scholar4. La Vecchia C, Pagano R, Negri E, Decarli A. Determinants of alcohol consumption in Italy [Letter]. Int J Epidemiol. 1987;16:295-6. [PMID: 3610459] CrossrefMedlineGoogle Scholar5. Ferraroni M, Decarli A, Franceschi S, La Vecchia C, Enard L, Negri E, . Validity and reproducibility of alcohol consumption in Italy. Int J Epidemiol. 1996;25:775-82. [PMID: 8921456] CrossrefMedlineGoogle Scholar Author, Article, and Disclosure InformationAuthors: Silvano Gallus, ScD; Silvia Franceschi, MD; Carlo La Vecchia, MDAffiliations: From Istituto di Ricerche Farmacologiche Mario Negri; 20157 Milan, Italy; International Agency for Research on Cancer; 69372 Lyon Cedex, France; and Istituto di Statistica Medica e Biometria, Universit degli Studi di Milano; 20133 Milan, Italy.Grant Support: By the Italian Association for Cancer Research, Milan, Italy. PreviousarticleNextarticle Advertisement FiguresReferencesRelatedDetailsSee AlsoUse of Postmenopausal Hormones, Alcohol, and Risk for Invasive Breast Cancer Wendy Y. Chen , Graham A. Colditz , Bernard Rosner , Susan E. Hankinson , David J. Hunter , JoAnn E. Manson , Meir J. Stampfer , Walter C. Willett , and Frank E. Speizer Metrics Cited byHormone therapy and breast cancer: what factors modify the association? 7 October 2003Volume 139, Issue 7Page: 601-602KeywordsAlcohol consumptionAlcoholsBreast cancerCohort studiesEthanolHabitsHormonesMenopauseOdds ratioWine ePublished: 7 October 2003 Issue Published: 7 October 2003 Copyright & PermissionsCopyright © 2003 by American College of Physicians. All Rights Reserved.PDF downloadLoading ...

aDepartment of Epidemiology, Istituto di Ricerche Farmacologiche Mario Negri bOccupational Health, Universita` di Milano, Italy Correspondence to Professor Carlo La Vecchia, MD, Istituto Mario Negri, Via G. La Masa, 19, Milano 20156, Italy Tel: +39 02 3901 4527; fax: +39 02 3320 02321; e-mail: [email protected] Received June 20, 2011 Accepted June 24, 2011

Changes in cigarette price have had an appreciable impact on smoking consumption in several countries. We analysed the price elasticity of demand for cigarettes in Italy over the period 1970-2001. A 1% increase in price of cigarettes led to a 0.30% decline in smoking prevalence and to a 0.43% decline in cigarette consumption. The present analyses confirm the existence of an inverse association between price and prevalence/consumption of cigarettes in Italy, and indicate that economic aspects may have important public health implications in tobacco control.