[Estimate of the population at risk between 2 census records for the calculation of incidence level or mortality of cancer: comparison of 4 methods].

The estimation of incidence (or mortality) rates for a given disease in a defined population needs precise knowledge of the population at risk of contracting the disease (or dying) during the study period. This article proposes and compares four methods for estimating population by age between two census surveys. The three methods which consist in aging the generations give results which are extremely close for age specific and standardized rates; on the contrary, the methods which do not take into account the generations at birth are inadequate due to the lack of births during the two world wars.