Biomarkers fulfil a crucial role in studying disease etiology, both with re
spect to biological plausibility, laboratory methodology, and in human stud
ies. Markers should preferably reflect the mainstream causal process releva
nt to a major proportion of disease. Such markers may provide the evidence
required to justify health claims regarding the role of antioxidants in the
etiology of chronic diseases, e.g., cancer, cardiovascular disease and cat
aract. Regarding study design, type A claims, addressing enhanced biologica
l function, would at least require human intervention studies. Type B claim
s, addressing reduced disease risk, would require prospective data, if poss
ible experimental. (Nested) case control studies may be used to establish t
he predictive validity of precursor lesions of disease that can serve to su
bstitute studies assessing disease occurrence as endpoints. Scientifically
sound and practically feasible requirements for justification of health cla
ims need be developed further.