[Epidemiologic criteria for the evaluation of positive and negative effects of the administration of high doses of vitamins].
Epidemiological methods can provide evidence, not proof, of cause effect relationships. They can confirm and quantify the existence of statistical relationships. Causality in epidemiological terms means that such relationships are strong and stable enough to enable the change of the suspected effect by changing the suspected cause. Criteria for epidemiological evidence of effects of elevated dosages of vitamins are basically the same as those for the evidence of effects of other exposures. Given the unambiguous classifications of both exposure and cases, they comprise strength, significance, specificity and consistency of the statistical association, plausible time relationships as well as dose effect relationship and consistency with other evidence. Epidemiological evidence for beneficial or detrimental effect of elevated vitamin doses may be derived from experimental or observational studies. It may confirm effects extrapolated from much higher exposure or from animal models. Today the term epidemiological evidence usually refers to field experience often to "observational", i.e. non-experimental evidence. Here comparison groups may not come from the same source populations and many confounding variables can only be controlled by sophisticated statistical procedures. A typical example for this are the so-called "ecological studies". They are frequently criticized because of their potential for exaggerated interpretations but they can be very helpful in constructing and supporting hypotheses. For very rare, and long term effects the description and evaluation of individual cases are often combined with attempts of quantification by relating them to the estimated exposure of the source population. This is subject to numerous sources of errors.(ABSTRACT TRUNCATED AT 250 WORDS)