Potential to Link Dietary Patterns in the Food Supply and Populations to Health

Extract

The article by Arem et al. ( 1 ) on the association of dietary patterns as defined by the Healthy Eating Index (HEI) – 2005 and pancreatic cancer contributes to a rapidly evolving literature on the association of dietary patterns with multiple disease outcomes. In summarizing the global literature on diet and cancer, the 2007 World Cancer Research Fund report concluded that the limited number of studies on dietary patterns and cancer outcomes prevented any conclusions on this topic and identified the need for more research to better define dietary patterns and examine their potential association with cancer ( 2 ). The study by Arem et al. is a well-designed and well-executed study that reflects efforts to advance research on dietary patterns. Such research is intended to address issues that have been well documented in nutrient or individual food or food group research, and these issues are clearly detailed by Arem and colleagues in their introduction: difficulty in disentangling the multicollinearity of nutrients and individual foods; inability to capture biologic interactions among different nutrients, foods, or nonnutritive components of diet; and improved distinction between individuals in terms of overall dietary patterns and quality at extremes of the distribution of a dietary pattern measure. In addition, just as foods and nutrients have been documented to be highly correlated, other health behaviors, such as energy intake and weight, alcohol intake, physical activity, and tobacco use, have been documented to be correlated with dietary patterns ( 3 ), as clearly demonstrated in the examination of these factors across the quintiles of HEI in Table 2 of the Arem et al. article ( 1 ). Nearly all of these health behaviors have been associated with risk of pancreatic cancer to some degree in cohort studies ( 4 ).