In this paper age–space–time models based on one and two-dimensional P-splines with B-spline bases are proposed for smoothing mortality rates, where both fixed relative scale and scale invariant two-dimensional penalties are examined. Model fitting and inference are carried out using integrated nested Laplace approximations, a recent Bayesian technique that speeds up computations compared to McMC methods. The models will be illustrated with Spanish breast cancer mortality data during the period 1985–2010, where a general decline in breast cancer mortality has been observed in Spanish provinces in the last decades. The results reveal that mortality rates for the oldest age groups do not decrease in all provinces.
MacNabYCDeanCB. Autoregressive spatial smoothing and temporal spline smoothing for mapping rates. Biometrics2001; 57: 949–956.
5.
UgarteMDGoicoaTMilitinoAF. Spatio-temporal modeling of mortality risks using penalized splines. Environmetrics2010; 21: 270–289.
6.
UgarteMDGoicoaTEtxeberriaJet al.A P-spline ANOVA-type model in space-time disease mapping. Stoch Env Res Risk A2012; 26: 835–845.
7.
BreslowNEClaytonDG. Approximate inference in generalized linear mixed models. J Am Stat Assoc1993; 88: 9–25.
8.
MacNabYCGustafsonP. Regression B-spline smoothing in Bayesian disease mapping: with an application to patient safety surveillance. Stat Med2007; 26: 4455–4474.
9.
MacNabYC. Spline smoothing in Bayesian disease mapping. Environmetrics2007; 18: 727–744.
10.
BauerCWakefieldJRueHet al.Bayesian penalized spline models for the analysis of spatio-temporal count data. Stat Med2016; 35: 1848–1865.
11.
SchrödleBHeldL. A primer on disease mapping and ecological regression using INLA. Computat Stat2011; 26: 241–258.
12.
RueHMartinoSChopinN. Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations. J Roy Stat Soc B2009; 71: 319–392.
13.
SunDTsutakawaRKKimHet al.Spatio-temporal interaction with disease mapping. Stat Med2000; 19: 2015–2035.
14.
ZhangSSunDHeCZet al.A Bayesian semi-parametric model for colorectal cancer incidences. Stat Med2006; 25: 285–309.
15.
PapoilaALRieblerAAmaral-TurkmanAet al.Stomach cancer incidence in Southern Portugal 1998-2006: a spatio-temporal analysis. Biometric J2014; 56: 403–415.
16.
EtxeberríaJGoicoaTLópez-AbenteGet al.Spatial gender-age-period-cohort analysis of pancreatic cancer mortality in Spain (1990–2013). Plos One2017; 12: 1–12.
17.
GoicoaTUgarteMDEtxeberríaJet al.Age–space–time CAR models in Bayesian disease mapping. Stat Med2016; 35: 2391–2405.
18.
International Organization Health. ICD-10a. International statistical classification of diseases and related health problems, http://www.who.int/classifications/icd/en/ (accessed 10 August 2017).
19.
FerlayJSteliarova-FoucherELortet-TieulentJet al.Cancer incidence and mortality patterns in Europe: estimates for 40 countries in 2012. Eur J Cancer2013; 49: 1374–1403.
20.
PollánMGarcía-MendizábalMJGómezBPet al.Situación epidemiológica del cáncer de mama en España. Psicooncología2007; 4: 231–248.
21.
BothaJLBrayFSankilaRet al.Breast cancer incidence and mortality trends in 16 European countries. Eur J Cancer2003; 39: 1718–1729.
22.
AscunceNSalasDZubizarretaRet al.Cancer screening in Spain. Ann Oncol2010; 21: 43–51.
23.
Álvaro-MecaADebónAGil-PrietoRet al.Breast cancer mortality in Spain: has it really declined for all age groups?Public Health2012; 126: 891–895.
24.
UgarteMDGoicoaTEtxeberriaJet al.Age-specific spatio-temporal patterns of female breast cancer mortality in Spain (1975–2005). Ann Epidemiol2010; 20: 906–9016.
25.
UgarteMDAdinAGoicoaT. One-dimensional, two-dimensional, and three dimensional B-splines to specify space-time interactions in Bayesian disease mapping: model fitting and model identifiability. Spatial Statistics2017. (in press), https://dx-doi-org.web.bisu.edu.cn/10.1016/j.spasta.2017.04.002.
26.
LerouxBGLeiXBreslowNEstimation of disease rates in small areas: a new mixed model for spatial dependence. In: HalloranMBerryD (eds). Statistical models in epidemiology, the environment and clinical trials, New York: Springer-Verlag, 1999, pp. 179–192.
27.
RueHHeldL. Gaussian Markov random fields: theory and applications, Boca Raton, FL: Chapman & Hall/CRC, 2005, pp. 95–110.
28.
WoodSNScheiplFFarawayJJ. Straightforward intermediate rank tensor product smoothing in mixed models. Stat Comput2013; 23: 341–360.
29.
SchrödleBHeldLRieblerAet al.Using integrated nested Laplace approximations for the evaluation of veterinary surveillance data from Switzerland: a case-study. J Roy Stat Soc C2011; 60: 261–279.
30.
SchrödleBHeldL. Spatio-temporal disease mapping using INLA. Environmetrics2011; 22: 725–734.
31.
UgarteMDAdinAGoicoaTet al.On fitting spatio-temporal disease mapping models using approximate Bayesian inference. Stat Meth Med Res2014; 23: 507–530.
32.
R Core Team. R: A language and environment for statistical computing., Vienna, Austria: R Foundation for Statistical Computing, http://www.R-project.org/2015 (accessed 10 August 2017).
33.
Martino S and Rue H. Implementing approximate Bayesian inference using integrated nested Laplace approximation: a manual for the INLA program. Department of Mathematical Sciences, NTNU, Norway, https://folk.ntnu.no/hrue/GMRFsim/manual.pdf (2010, accessed 10 August 2017).
34.
GoicoaTAdinAUgarteMDet al.In spatio-temporal disease mapping models, identifiability constraints affect PQL and INLA results. Stochastic Environmental Research and Risk Assessment2017. DOI: 10.1007/s00477-017-1405-0.
35.
GelmanA. Prior distributions for variance parameters in hierarchical models (comment on article by Browne and Draper). Bayesian Analys2006; 1: 515–534.
36.
CarrollRLawsonABFaesCet al.Comparing INLA and OpenBUGS for hierarchical Poisson modeling in disease mapping. Spatial Spatio-temporal Epidemiol2015; 14–15: 45–54.
37.
SimpsonDRueHRieblerAet al.Penalising model component complexity: a principled, practical approach to constructing priors. Stat Sci2017; 32: 1–28.
38.
SpiegelhalterDJBestNGCarlinBPet al.Bayesian measures of model complexity and fit. J Roy Stat Soc B2002; 64: 583–639.
39.
PlummerM. Penalized loss functions for Bayesian model comparison. Biostatistics2008; 9: 523–539.
40.
WatanabeS. Asymptotic equivalence of Bayes cross validation and widely applicable information criterion in singular learning theory. J Mach Learn Res2010; 11: 3571–3594.
41.
GneitingTRafteryAE. Strictly proper scoring rules, prediction, and estimation. J Am Stat Assoc2007; 102: 359–378.
42.
CurrieIDDurbánMEilersPH. Smoothing and forecasting mortality rates. Stat Model2004; 4: 279–298.
43.
CurrieIDDurbánMEilersPH. Generalized linear array models with applications to multidimensional smoothing. J Roy Stat Soc B2006; 68: 1–22.
44.
Wood SN. Mixed GAM computation vehicle with GCV/AIC/REML smoothness estimation 2017. R package version 1.8–17.
45.
ErbasBAkramMGertigDMet al.Using functional data analysis models to estimate future time trends in age-specific breast cancer mortality for the United States and England-Wales. J Epidemiol2010; 20: 159–165.
46.
HolfordTRCroninKAMariottoABet al.Changing patterns in breast cancer incidence trends. J Natl Cancer Inst Monogr2006; 36: 19–25.
47.
ClèriesRRibesJEstebanLet al.Time trends of breast cancer mortality in Spain during the period 1977–2001 and Bayesian approach for projections during 2002–2016. Ann Oncol2006; 17: 1783–1791.
48.
RenardFVan EyckenLArbynM. High burden of breast cancer in Belgium: recent trends in incidence (1999–2006) and historical trends in mortality (1954–2006). Arch Public Health2011; 69: 2–2.
