Nutritional Epidemiology in Practice

Because randomized controlled trials that track whole diets over the decades needed to observe chronic disease outcomes are costly and often impractical, nutritional epidemiology relies heavily on observational study designs, including large prospective cohort studies, case-control studies, and cross-sectional studies. Dietary exposure in these studies is most often estimated using food frequency questionnaires (FFQs), 24-hour dietary recalls, or food diaries, all of which are subject to measurement error and recall bias, since they depend on participants accurately remembering and reporting what they ate. Doubly labeled water is considered a gold-standard, objective method for measuring total energy expenditure, but it is rarely used to assess the composition of what people actually eat. Mendelian randomization, a technique that uses genetic variants associated with a particular exposure as a proxy to help infer causal relationships and reduce the influence of confounding, has become an increasingly important complementary tool for strengthening causal inference in nutrition research.