Introduction
Only the first question posed to find the 5 star antioxidant, (see bioavailability) can be answered using biochemical analytical in vitro methods. Any further questions require in vivo biochemical/physiological assays to provide the answers. The transfer through the gut of phytonutrients from the digested food into the bloodstream is called absorption, and pre- and post-prandial sampling of the human subject's blood for the target compound requires stringent ethical sanctions and statistically accurate and reproducible clinical procedures to ensure the data are meaningful and verifiable by peer validation. This normally means that a number of volunteer subjects are required. They are standardised before the experiment starts. They are then divided into at least two groups, one of which receives a placebo. Depending on the objectives, other physiological samples may be taken e.g. urine, and a number of timed blood samples taken to monitor the post prandial state, to determine the duration of the antioxidant effect.

Methods of Analysis
Ideally, dealing with variable human subjects, a statistically rigorous method would be chosen, such as, the double-blind placebo-controlled trial, so that neither the subjects nor the scientists running the trial know which group is receiving the active ingredient. This is not always possible and compromises have to be made. There are a great number of observational methods and therefore we will only mention those commonly used to measure dietary effects.
1. Cohort studies
All the relevant physical data e.g. blood pressure, of a similar aged group population, along with their habits of interest such as diet and exercise regimes are recorded at the beginning of the trial and the group followed over time to ascertain the consequences of dietary intervention on health and well being, to an end-point. Cohort studies are preferred, but data if it wasn't thought to be needed at the beginning of the trial cannot be added retrospectively.
2. Case-control Studies
Researchers identify a group of people for study and choose another set of people to act as controls. For example the case subjects may be on a dairy-free diet and the control set on a diet containing dairy produce.
3. The nested case-control study
The characteristics of people on cohort studies are known and they can participate in additional case-control studies. This is a nested case-control study, and is a powerful method.
4. The meta-analysis
Is a statistical analysis of a large collection of similar studies. The recent meta study for the World Cancer Research Fund identified 500,000 relevant studies and selected 7,000 for further analysis, from which the conclusions were drawn. However, the preliminary selection groups who reduced the number to 7,000, if constituted differently, may well have chosen a different set of papers for the final analysis, and come to a different conclusion.