Incubation periods of cancer: old and new.

Following a review of the early efforts in occupational studies and animal experiments that were useful to elucidate concepts of latency and incubation periods in cancer, the paper presents the problem of definitions. In addition to semantic issues, conceptual models have influenced the development of appropriate empirical research in this field. The problems of measurement facing studies of incubation periods include; the assessment of time and nature of exposure, the definition of disease onset and the variation of sources of data used in the estimation of these periods. The different approaches that have been used to measure the incubation periods of cancer include; the direct approach, a study of changes of relative risks with time, calculation of time sequence between an outbreak of disease and the introduction of a suspected exposure, population based incidence studies, analysis of anecdotal information and studies of case clusters using different methods of establishing case linkages. There are well defined practical applications for the study of incubation periods of cancer. These are primarily in the area of etiological investigation, or may have practical relevance in community and patient management in past and future risk assessment. In its final part the paper raises issues that are researcheable in this subject. Does the incubation period of a cancer vary with the different etiologies in a multifactorial model? What is the effect of the dose of carcinogen on the length of the incubation period? Does the incubation period vary with age?