A disease can be recognised as an occupational disease, if a connection between a specific occupation and a specific disease is "more likely than not". The concept of "risk doubling" (which is the same as a relative risk of 2) at a first glance constitutes the direct translation of this social law concept into the epidemiological terminology. A closer look, however, reveals that the concept of the "relative risk" consists of two different benchmarks - the cumulative risk ratio CRR and the incidence rate ratio IRR. But only the CRR allows a direct derivation of a statement about the causation probability. As this paper shows at the example of a hypothetical cohort, it would, however, be necessary to differentiate by age to answer the "dependency question" on the basis of the CRR. Furthermore, it must be considered that the negative effects of an occupational exposure does not only show in the occurrence of a disease which would not have occurred without the exposure; an occupational exposure may as well lead to an acceleration of the onset of a disease that would have occurred later without exposure. This temporal influence on a disease's onset is not reflected by the CRR at all; this makes the CRR insufficient as the sole parameter for the determination of causation. With the IRR it is - as a rule - possible to gain occupational risk estimates that are not dependent on age, but as it considers not only the "excess risk", but also the time of occurrence an interpretation of the IRR with regard to the causation probability is quite difficult. The practical use of the alternative method of "risk and rate advancement periods" (RAP) for the quantification of a temporal back shift (acceleration) of diseases is discussed in this paper on the basis of a case control study concerning the risk factors of osteoarthrosis of the knee. The RAP does not allow a direct interpretation with regard to the causation probability. Therefore, if the RAP shall be used to answer the "occupational diseases question", a convention would be required to clarify whether the acceleration of the occurrence risk of a specific disease by 5, 10 or 20 years fulfils the legal recognition criteria. The number of years considered by social law (the lengths of the RAP) should of necessity depend on the severity of the disease in question. In order to initiate and support the required discussion with regard to disease-specific lengths of legally "essential" RAPs for diseases that are clearly age-dependent, the authors would like to encourage the consideration of both relative risk parameters (e.g. CRR, IRR) and acceleration parameters (e.g. RAP) for the presentation of results of epidemiological studies.