RESUMEN
Besides the Mediterranean diet, there is a paucity of studies examining plant-based diets in relation to cancer outcomes in Mediterranean populations. We analyzed 22,081 apparently cancer-free participants (mean age 55 ± 12 year) from the Moli-sani study (enrollment period 2005-2010; Italy). A general pro-vegetarian food pattern was computed by assigning positive or negative scores to plant- or animal-derived foods, respectively from a 188-item FFQ. A priori healthful or unhealthful pro-vegetarian food patterns distinguished between healthy plant foods (e.g., fruits, vegetables) and less-healthy plant foods (e.g., fruit juices, refined grains). Cancer incidence was defined as the earliest diagnosis of cancer from hospital discharge records over a median follow-up of 12.9 years. In multivariable-adjusted analyses, a general pro-vegetarian food pattern was associated with a lower rate of cancer incidence (HR = 0.85; 95%CI 0.75-0.97 for Q5 vs. Q1); no association was observed between the healthful or unhealthful pro-vegetarian food patterns and overall cancer incidence. A healthful pro-vegetarian pattern, however, was inversely associated with digestive cancer (HR = 0.76; 95%CI 0.58-0.99 for Q5 vs. Q1), while the unhealthful pro-vegetarian pattern was directly linked to respiratory cancer (HR = 1.68; 95%CI 1.06-2.68 for Q5 vs. Q1). Our findings in a Mediterranean population support the hypothesis that some, but not all pro-vegetarian diets, might prevent some cancers.