Bayesian Bivariate Disease Mapping

  • Richard G. Feltbower
  • Samuel O. M. Manda


There has been substantial progress in the development of Bayesian spatial modelling and estimation in recent years to overcome the problem of the sparseness of data across small geographical areas for rare diseases. Attention has also focused on developing spatial models to accommodate multivariate disease mapping, for example in situations where one wishes to test common epidemiological or aetiological features among different conditions. This chapter expands on this work by comparing a classical frequentist approach to full Bayesian estimation for fitting a bivariate spatial disease model. As an illustration, we apply the models to population-based childhood leukaemia and childhood diabetes data from Yorkshire, United Kingdom to determine the similarity in their spatial distribution. The spatial association between the two diseases is modelled using a multivariate normal distribution on the spatial and heterogeneity components within a hierarchical Bayesian random effects model. The effect on the degree of spatial correlation after adjusting for socio-demographic factors previously associated with disease incidence is also assessed.


Markov Chain Monte Carlo Acute Lymphoblastic Leukaemia Deviance Information Criterion Spatial Random Effect Markov Chain Monte Carlo Technique 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.



We are grateful to Oxford University Press for permission to reproduce tables which originally appeared in the following article: American Journal of Epidemiology, Vol 161, Issue 11 , pp 1168–1180, 2005, “Detecting small-area similarities in the epidemiology of childhood acute lymphoblastic leukemia and diabetes mellitus, Type 1: a Bayesian approach.”


  1. Besag, J., York, J., & Mollie, A. (1991). Bayesian image restoration, with two applications in spatial statistics (with discussion). Annals of the Institute of Statistical Mathematics, 43, 1–75.CrossRefGoogle Scholar
  2. Clayton, D., & Kaldor, J. (1987). Empirical Bayes estimates of age-standardized relative risks for use in disease mapping. Biometrics, 43, 671–681.PubMedCrossRefGoogle Scholar
  3. Congdon, P. (2003). Applied Bayesian models. Chichester: Wiley.CrossRefGoogle Scholar
  4. Dabney, A. R., & Wakefield, J. C. (2005). Issues in the mapping of two diseases. Statistical Methods in Medical Research, 14, 83–112.PubMedCrossRefGoogle Scholar
  5. Feltbower, R. G., McKinney, P. A., Parslow, R. C., Stephenson, C. R., & Bodansky, H. J. (2003). Type 1 diabetes in Yorkshire, UK: Time trends in 0–14 and 15–29 year olds, age at onset and age-period-cohort modelling. Diabetic Medicine, 20, 437–441.PubMedCrossRefGoogle Scholar
  6. Feltbower, R. G., McKinney, P. A., Greaves, M. F., Parslow, R. C., & Bodansky, H. J. (2004). International parallels in leukemia and diabetes epidemiology. Archives of Disease in Childhood, 89, 54–56.PubMedCrossRefGoogle Scholar
  7. Feltbower, R. G., Manda, S. O. M., Gilthorpe, M. S., Greaves, M. F., Parslow, R. C., Kinsey, S. E., Bodansky, H. J., & McKinney, P. A. (2005). Detecting small-area similarities in the epidemiology of childhood acute lymphoblastic leukemia and diabetes mellitus, type 1: A Bayesian approach. American Journal of Epidemiology, 161, 1168–1180.PubMedCrossRefGoogle Scholar
  8. Gelman, A., & Rubin, D. B. (1992). Inference from iterative simulations using multiple sequences. Statistical Science, 7, 457–472.CrossRefGoogle Scholar
  9. Gilks, W. R., Richardson, S., & Spiegelhalter, D. J. (1996). Markov Chain Monte Carlo in practice. London: Chapman and Hall.Google Scholar
  10. Greaves, M. (1997). Etiology of acute leukemia. The Lancet, 349, 344–349.CrossRefGoogle Scholar
  11. Held, L., Natario, I., Fenton, S. E., Rue, H., & Becker, N. (2006). Towards joint disease mapping. Statistical Methods in Medical Research, 14, 61–82.CrossRefGoogle Scholar
  12. Jin, X., Carlin, B. P., & Banerjee, S. (2005). Generalised hierarchical multivariate CAR models for areal data. Biometrics, 6, 539–557.Google Scholar
  13. Knorr-Held, L., & Best, N. G. (2001). A shared component model for detecting joint and selective clustering of two diseases. Journal of the Royal Statistical Society A, 164, 73–85.CrossRefGoogle Scholar
  14. Langford, I. H., Leyland, A. H., Rasbash, J., & Goldstein, H. (1999). Multilevel modelling of the geographical distributions of diseases. Journal of the Royal Statistical Society C, 48, 253–268.CrossRefGoogle Scholar
  15. Lawson, A. B. (2008). Bayesian disease mapping: Hierarchical modeling in spatial epidemiology. Boca Raton: CRC Press.CrossRefGoogle Scholar
  16. Lawson, A. B., Browne, W. J., & Vidal Rodeiro, C. L. (2003). Disease mapping with WinBUGS and MLwiN. London: Wiley.CrossRefGoogle Scholar
  17. Leyland, A. H., Langford, I. H., Rabash, J., & Goldstein, H. (2000). Multivariate spatial models for event data. Statistics in Medicine, 19, 2469–2478.PubMedCrossRefGoogle Scholar
  18. Lunn, D. J., Thomas, A., Best, N., & Spiegelhalter, D. (2000). WinBUGS – A Bayesian modelling framework: Concepts, structure, and extensibility. Statistics and Computing, 10, 325–337.CrossRefGoogle Scholar
  19. Mardia, K. V. (1988). Multi-dimensional multivariate Gaussian Markov random fields with application to image processing. Journal of Multivariate Analysis, 24, 265–284.CrossRefGoogle Scholar
  20. McCullagh, P., & Nelder, J. A. (1989). Generalized linear models (2nd ed.). London: Chapman and Hall.Google Scholar
  21. McKinney, P. A., Parslow, R. C., Lane, S. A., Lewis, I. J., Picton, S., Kinsey, S. E., & Bailey, C. C. (1998). Epidemiology of childhood brain tumors in Yorkshire, UK 1974–1995: Changing patterns of occurrence. British Journal of Cancer, 78, 974–979.PubMedCrossRefGoogle Scholar
  22. Parslow, R. C., McKinney, P. A., Law, G. R., & Bodansky, H. J. (2001). Population mixing and childhood diabetes. International Journal of Epidemiology, 30, 533–538.PubMedCrossRefGoogle Scholar
  23. Smith, A. F. M. (1993). Bayesian computations via the Gibbs sampler and related Markov Chain Monte Carlo methods. Journal of the Royal Statistical Society B, 55, 3–23.Google Scholar
  24. Stiller, C. A., & Boyle, P. J. (1996). Effect of population mixing and socioeconomic status in England and Wales, 1979–85, on lymphoblastic leukemia in children. BMJ, 313, 1297–1300.PubMedCrossRefGoogle Scholar
  25. Townsend, P., Phillimore, P., & Beattie, A. (1988). Health and deprivation: Inequality and the North. London: Croom Helm.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2012

Authors and Affiliations

  1. 1.Division of Epidemiology, Centre for Epidemiology and Biostatistics, Leeds Institute of Genetics, Health & TherapeuticsUniversity of LeedsLeedsUK
  2. 2.Biostatistics UnitSouth Africa Medical Research CouncilPretoriaSouth Africa

Personalised recommendations