It is important to realize that there will almost certainly never be controlled trials of meat consumption (or any other single food) and chronic disease endpoints because of the cost, size of the study sample needed, length of time required, as well as the issue of compliance with dietary advice in both the intervention and control groups. There is also the ethical issue that if meat is considered a risk factor for any disease endpoint based on interpretation of existing data, then it would not be ethical to feed one group of subjects with an amount of meat that would be expected to lead to increased disease. Therefore, most of the data available on which to base dietary recommendations is derived from observational studies of free-living people. It must be acknowledged that cross-sectional studies, which are done at a single time point, do not have the ability to relate cause and effect by themselves. This position was articulated by Hill 50 years ago, who listed nine types of evidence that would help bolster findings from observational studies, and remains well regarded(Hill, 1965); the nine types of evidence include: strength which Hill believed was by far the most important, consistency, specificity, temporality, biological gradient, plausibility, coherence, experiment, and analogy. These considerations are lauded simultaneously to support