Age–period–cohort models and disease mapping

Abstract

SUMMARYJoint modelling of space and time variation of the risk of disease is an important topic in descriptive epidemiology. Most of the proposals in this field deal with at most two time scales (age-period or age-cohort). We propose a hierarchical Bayesian model that can be used as a general framework to jointly study the evolution in time and the spatial pattern of the risk of disease. The rates are modelled as a function of purely spatial terms (local effects of risk factors that do not vary in time), time effects (on the three time axes: age, calendar period and birth cohort) and space-time interactions that describe area specific time patterns.

A space–time analysis of the relationship between material deprivation and mortality for lung cancer

“…Table 1 reports observed age and period specific rates (Â 100 000) and number of cases. For a space-time analysis of a subset of these data (male lung cancer in Tuscany from 1971 to 1994), that considers ageperiod and cohort effects; see Lagazio et al (2002). Here we focus only on period effects.…”

Spatio‐temporal disease mapping using INLA

“…This approach was applied to a number of different models, e.g. to age-period-cohort models (Lagazio et al, 2003). Recently, an alternative approach has been proposed by Martínez-Beneito et al (2008) that offers an autoregressive approach to disease mapping by adapting ideas from autoregressive time series and spatial modeling to link information in time and space, respectively.…”

“…This class of models is usually formulated within a hierarchical Bayesian framework with latent Gaussian model (Besag et al, 1991;Banerjee et al, 2004). Several proposals have been made including a parametric (Bernardinelli et al, 1995b;Assunção et al, 2001) and nonparametric (Knorr-Held, 2000;Lagazio et al, 2003;Schmid and Held, 2004) formulation of the time trend and the respective space-time interactions. To obtain the respective parameter estimates computationally expensive Markov chain Monte Carlo (MCMC) algorithms are typically used, which might induce a large Monte Carlo error of the parameter estimates.…”

Real geographies and virtual landscapes: Exploring the influence on place and space on mortality Lexis surfaces using shaded contour maps

“…There are also a number of existing age-period-cohort (APC) models, some of which explicitly include spatial factors (for example: Lagazio et al 2003;Aamodt et al 2007;Xu and Hertzberg 2013). It may be interesting to compare estimates produced by these models with the imputed surface approach described above, as well as with estimates produced informally, through visual inspection of the contour lines.…”