Design considerations on the proof of efficacy of functional foods

Functional food research encompasses several types of study designs, including observational studies and randomized clinical trials (RCTs). Markers that can predict potential benefits or risks relating to certain health conditions are often the primary endpoints of such studies since a direct measurement of the effect of food on health and well-being and/or reduction of disease risk is often not possible. Whether RCT should be at the top of the pyramid also in nutritional research remains a controversial issue. Undoubtedly, further research is needed to redesign RCT methodology that would adequately serve the need to demonstrate the health effects of foods. We address this functional food research question by assuming that there is a known relationship between the surrogate and the true endpoint explored during the surrogacy assessment. Statistical inference on the true (unobserved) endpoint is derived on the basis of its predicted values. We illustrate this approach through a motivating example from the literature on coronary heart disease, integrated with simulated scenarios.