Methods to study the deprivation and its relationships with cancer incidence in a local area [Metodi per lo studio della deprivazione e le sue relazioni con l’incidenza del cancro in un’area locale].

Abstract Aim: This study was aimed at finding the best method to describe the relationships between cancer incidence and deprivation of population at a local level (the Genoa province). Methods: The population was clustered using the National Deprivation Index (NDI) and the Genoa Deprivation Index (GDI), both using data drawn from the 2001 Census, either at municipality or at district levels. The NDI and GDI used were drawn from the NDI 2001 and the Liguria Region Deprivation Index 2001 (LRDI), re-parameterised on the values of areas pertaining to the Genoa province considering the quintiles of the population. In the building of both original indexes five traits representing the multi-dimensionality of the social and material deprivation concept were selected, but each one resulted composed by quite opposite variables. The derived indexes were applied to clustering Genoa province population computing for each cluster the cancer incidence occurrence in the period 1999-2003. For the five deprivation clusters of population, individuated aggregating the 91 areas of the province on the basis of the two indexes, the Standardised Incidence Ratios (SIR) were computed for all malignant cancers combined (skin carcinomas excluded) and 35 sites. The observed SIR and their 95% confidence intervals were calculated by gender and age (all ages, 0-64 years and 65+ years). To evaluate if deprivation influenced the incidence, an analysis was conducted between SIR group variances. Results: From the application of the GDI, many significant associations between cancer incidence and deprivation surfaced, particularly for women. However, the NDI did not correctly identify the main part of the negative linear trends whereas the GDI did so. The lacking of statistical significance in all cancers combined distribution by GDI in men was mainly explained by the opposite trends in lung and prostate cancers, while in women the breast negative trend was balanced by a major number of different sites. Conclusions: Although the GDI cannot be considered a “pure” deprivation index, it does seem to be useful for public health purposes due to its capability of discriminating SE clusters in a fairly homogenous population, like the Liguria one.