WASHINGTON—Size matters, especially in epidemiologic studies. The European Prospective Investigations into Cancer and Nutrition (EPIC) was founded on that principle in 1992, said Elio Riboli, MD, ScM, MPH.

Since then, the EPIC researchers have recruited 484,000 subjects from nine European countries for a multicenter, prospective study of the relationship between dietary, genetic, metabolic, and lifestyle characteristics and the risk of cancer and cardiovascular disease.

Dr. Riboli, chief of the Nutrition and Cancer Unit of the International Agency for Research on Cancer, Lyon, France, described the research project at the 10th Annual Research Conference of the American Institute for Cancer Research (AICR).

“We know that a plant-based diet protects against cancers of the colorectum, stomach, upper aerodigestive tract, and possibly others,” Dr. Riboli said. “We also know that salty foods raise the risk of stomach cancer and that there is information suggesting that eating meat increases colorectal cancer risk. But we need to know more: Which foods? Which vegetables? What is their chemical and physical structure?”

In short, while “eat more fruits and vegetables” may be good advice for the public, more detailed information is needed to determine the role of nutrition in initiating or preventing the development of many cancers, he said.

An advantage of studying populations across Europe (in Denmark, France, Germany, Greece, Italy, the Netherlands, Spain, Sweden, and the United Kingdom) is that wide variations in cancer incidence and dietary practices can be explored.

For instance, Dr. Riboli said, the breast cancer rate in Scotland is four times higher than the rate in Sicily or southern Spain, while the stomach cancer rate is highest in central Spain and northern Italy and lowest in England.

The EPIC researchers have collected a detailed questionnaire from each study subject on diet, lifestyle, illness, medication use, physical activity, alcohol and tobacco use, reproductive and contraceptive history, and socioeconomic factors. Anthropometric measurements have also been taken.

The food questionnaire lists 150 to 300 food choices, and a more detailed inquiry into a 1-day “actual” diet taken from 7% of the respondents will cover 3,000 possible food descriptions and 700 recipes tailored for each country. “At worst,” Dr. Riboli quipped, “we can publish books of recipes, if we fail as cancer researchers.”

To date, the EPIC group has taken blood samples from 387,000 subjects. Blood from each subject is parceled out into 28 small (0.5 mL) plastic straws containing plasma, serum, buffy coat or red blood cells, and stored in liquid nitrogen for study over the next 10 to 15 years. The storage system is designed to withstand power outages as long as 4 months without refilling.

EPIC will use these stored biologic samples to measure biomarkers that will then be correlated with the 17,000 cases of cancer expected to occur in this cohort by 2002. Prediagnostic levels of steroid hormones, insulin-like growth factors, and other metabolic factors will be correlated with inherited genetic susceptibilities.

“We are trying to go beyond the black box of epidemiology and identify mechanisms that lead to cancers,” Dr. Riboli said. This may lead to answers to current puzzles such as whether levels of eicosapentaenoic acid (EPA) from fish oil are related to cancer risk. EPA levels have been shown to be higher in Denmark and lower among a group of vegetarians in Oxford, UK.

These biomarker studies may also lead to improved measurements of data collected on dietary questionnaires, whose accuracy has long been debated by nutrition researchers.