DIET AND THYROID-CANCER - A POOLED ANALYSIS OF 4 EUROPEAN CASE-CONTROL STUDIES

The relationship between frequencies of consumption of selected indicator foods and the risk of thyroid cancer was investigated in a pooled analysis of 4 case-control studies conducted in 3 areas of northern Italy and the Swiss Canton of Vaud, on a total of 385 histologically confirmed cases of thyroid cancer and 798 controls in hospital for acute, non-neoplastic, non-hormone-related disease. Cases tended to consume significantly more frequently a number of starchy foods: the odds ratio (OR) for the highest vs. lowest tertile was 1.8 for pasta or rice, 2.1 for bread, 1.6 for pastry and 2.0 for potatoes. ORs also tended to be above unity for several types of meat and significantly so for chicken and poultry, cooked ham, salami and sausages. Raw ham and fish were significantly protective (OR = 0.7 in the highest tertile for both). Significant direct associations were observed with cheese (OR = 1.4 for the highest tertile), butter (OR = 2.1) and oils other than olive (OR = 1.6). The risk estimates were below unity for most types of vegetables and fruits, and the inverse trends were significant for carrots (OR = 0.6 for the highest tertile), green salad (OR = 0.6) and citrus fruits (OR = 0.7). No association was observed with alcohol intake. These results were consistent and reproducible across various study centers.