One of the aims of the EPIC study is to produce accurate descriptions of the dietary habits of the participants recruited in the 27 EPIC centers of 10 European countries. To do this, different dietary assessment instruments were developed and applied to capture the wide range of diets characterizing the different European populations. Three different food frequency questionnaires were developed for Italy: one for the centers of Varese, Turin and Florence, one for Ragusa and one for Naples. These inquired about eating habits over the previous year and were completed by 46,839 Italian EPIC participants. Specially developed software analyzed the responses and linked them to food composition tables in order to provide a nutritional breakdown of individual and collective diets. A further aim of EPIC was to develop a method of rendering data from different dietary questionnaires comparable. To do this, dietary data were collected from a sample of about 8% of the Italian EPIC cohort, using a standardized computer-driven 24-hour dietary recall interview, and then compared with the dietary data collected by the questionnaires. This paper provides an extensive description of the technical features and performance of the food frequency questionnaires and the 24-hour recall interview, including a comparison of estimates of the intake of different food groups provided by the two instruments. From this comparison, the repeatability and reliability of consumption estimates was assessed, resulting in indications for improving data comparability. The paper also presents food frequency questionnaire estimates of the daily intake of foods and nutrients by center, sex and age group, as well as information on dietary habits such as place and time of intake, and food preparation and preservation methods as provided by the 24-hour recall interview. The picture that emerged is that Italian eating habits are undergoing marked changes, with a tendency to less healthy eating. Documentation of these changes in relation to age, sex and region provides an essential starting point for investigating the influence of diet on the development of cancer and other chronic diseases.

Abstractbjective:The aim of this study was to compare the quantities of alcohol and types of alcoholic beverages consumed, and the timing of consumption, in centres participating in the European Prospective Investigation into Cancer and Nutrition (EPIC). These centres, in 10 European countries, are characterised by widely differing drinking habits and frequencies of alcohol-related diseases.Methods:We collected a single standardised 24-hour dietary recall per subject from a random sample of the EPIC cohort (36 900 persons initially and 35 955 after exclusion of subjects under 35 and over 74 years of age). This provided detailed information on the distribution of alcohol consumption during the day in relation to main meals, and was used to determine weekly consumption patterns. The crude and adjusted (by age, day of week and season) means of total ethanol consumption and consumption according to type of beverage were stratified by centre and sex.Results:Sex was a strong determinant of drinking patterns in all 10 countries. The highest total alcohol consumption was observed in the Spanish centres (San Sebastian, 41.4 g day−1) for men and in Danish centres (Copenhagen, 20.9 g day−1) for women. The lowest total alcohol intake was in the Swedish centres (Umeå, 10.2 g day−1) in men and in Greek women (3.4 g day−1). Among men, the main contributor to total alcohol intake was wine in Mediterranean countries and beer in the Dutch, German, Swedish and Danish centres. In most centres, the main source of alcohol for women was wine except for Murcia (Spain), where it was beer. Alcohol consumption, particularly by women, increased markedly during the weekend in nearly all centres. The German, Dutch, UK (general population) and Danish centres were characterised by the highest percentages of alcohol consumption outside mealtimes.Conclusions:The large variation in drinking patterns among the EPIC centres provides an opportunity to better understand the relationship between alcohol and alcohol-related diseases.

Background and Purpose-The association between obesity and atherosclerotic disease is controversial. In the present analysis, we evaluated whether common carotid intima-media thickness (IMT) and area, 2 markers of preclinical atherosclerosis, were increased in obese subjects. Methods-More than 5000 middle-aged women (nϭ5062; age, 30 to 69 years) living in the area of Naples, Southern Italy, were recruited for a prospective, currently ongoing study on the etiology of cardiovascular disease and cancer in the female population (the Progetto ATENA study). A subsample of 310 participants underwent high-resolution B-mode ultrasound examination, and the IMTs, intima-media areas, and lumen diameters of common carotid arteries were measured with a semiautomated computerized program. Subjects were divided into 3 groups on the basis of the recently published obesity guidelines for body mass index (BMI), a marker of general obesity, and tertiles of waist-to-hip ratio (WHR), a marker of regional obesity. Results-Women with a BMI Ն30 kg/m 2 showed higher systolic and diastolic blood pressures, triglycerides, and fasting glucose and insulin, as well as lower high-density lipoprotein concentrations, than subjects with lower BMI. A gradual increase in common carotid IMT and intima-media area was observed when lean women (0.94Ϯ0.01 mm and 19.8Ϯ0.5 mm 2 , respectively) were compared with overweight (0.98Ϯ0.01 mm and 21.0Ϯ0.4 mm 2 ) and obese (1.02Ϯ0.02 mm and 22.6Ϯ0.8 mm 2 , PϽ0.005 for linear trend) individuals. Similarly, women in the highest tertile of WHR (Ͼ0.85) had adverse risk factor profiles and thicker carotid intima-media complex than those in the first 2 tertiles (PϽ0.01 and PϽ0.05 for IMT and intima-media area, respectively). In multivariate analyses, BMI and WHR were significant predictors of carotid wall thickness, independently of other traditional and nontraditional cardiovascular risk factors (age, blood pressure, lipid abnormalities, fasting insulin). Conclusions-The present results indicate a graded and independent association between general and abdominal obesity-reflected by high BMI and WHR-and carotid artery wall thickening in a population of middle-aged women.

EPIC-Italy is the Italian section of a larger project known as EPIC (European Prospective Investigation into Cancer and Nutrition), a prospective study on diet and cancer carried out in 10 European countries. In the period 1993-1998, EPIC-Italy completed the recruitment of 47,749 volunteers (15,171 men, 32,578 women, aged 35-65 years) in 4 different areas covered by cancer registries: Varese (12,083 volunteers) and Turin (10,604) in the Northern part of the country; Florence (13,597) and Ragusa (6,403) in Central and Southern Italy, respectively. An associate center in Naples enrolled 5,062 women. Detailed information for each individual volunteer about diet and life-style habits, anthropometric measurements and a blood sample was collected, after signing an informed consent form. A food frequency questionnaire specifically developed for the Italian dietary pattern was tested in a pilot phase. A computerized data base with the dietary and life-style information of each participant was completed. Blood samples were processed in the same day of collection, aliquoted (RBC, WBC, serum and plasma) and stored in liquid nitrogen containers. Follow-up procedures were validated and implemented for the identification of newly diagnosed cancer cases. Cancer incidence was related to dietary habits and biochemical markers of food consumption and individual susceptibility in order to test the role of diet-related exposure in the etiology of cancer and its interaction with other environmental or genetic determinants. The comparability of information in a prospective study design is much higher than in other studies. The availability of such a large biological bank linked to individual data on dietary and life-style exposures also provides the unique opportunity of evaluating the role of selected genotypes involved in the metabolism of chemical compounds and DNA repair, potentially related to the risk of cancer, in residents of geographic areas of Italy characterized by specific cancer risk and different dietary patterns. Baseline characteristics of participants are briefly described.

Background: In addition to their possible direct biological effects, plasma carotenoids can be used as biochemical markers of fruit and vegetable consumption for identifying diet -disease associations in epidemiological studies. Few studies have compared levels of these carotenoids between countries in Europe. Objective: Our aim was to assess the variability of plasma carotenoid levels within the cohort of the European Prospective Investigation into Cancer and Nutrition (EPIC). Methods: Plasma levels of six carotenoids -a-carotene, b-carotene, b-cryptoxanthin, lycopene, lutein and zeaxanthin -were measured cross-sectionally in 3043 study subjects from 16 regions in nine European countries. We investigated the relative influence of gender, season, age, body mass index (BMI), alcohol intake and smoking status on plasma levels of the carotenoids.