PURPOSE OF REVIEW: Our purpose was to discuss the methodological limitations of observational nutritional epidemiology research, using observational studies on coffee intake and health as a case example.
RECENT FINDINGS: A number of recent observational studies on the potential health effects of daily coffee intake have reported protective associations between higher coffee intake and a variety of health outcomes, including death. This is inconsistent with the findings from classic studies showing an increased risk of coronary heart disease events, performed in young adults with a homogeneous education level, and adjusting for tobacco use.
SUMMARY: Many nutritional epidemiological studies have important limitations, which limit their validity. These include the use of prevalent user designs, risk of reverse causality, measurement error particularly of the exposure of interest, and residual confounding by socioeconomic status. In this review we discuss these potential issues, and provide constructive recommendations intended to help minimize them.
