Skip to main content

Advertisement

Log in

A flexible link for joint modelling longitudinal and survival data accounting for individual longitudinal heterogeneity

  • Original Paper
  • Published:
Statistical Methods & Applications Aims and scope Submit manuscript

Abstract

This work aims at jointly modelling longitudinal and survival HIV data by considering the sharing of a set of parameters of interest. For the CD4 longitudinal stochastic process we propose a regression model where individual heterogeneity is allowed to vary in terms of the mean and the variance, relaxing the usual assumption of a common variance for the longitudinal residuals. Along, we will be considering a hazard regression model to analyse the time between HIV/AIDS diagnostic and death. For introducing enough flexibility in the structure linking the longitudinal and survival processes, we consider time-varying coefficients. That is achieved using Penalized Splines and allows the relationship to vary in time. The CD4 residuals standard deviation is considered as a covariate in the hazard model, thus enabling to study the effect of the CD4 counts’ stability on the survival. The proposed framework surpasses the performance of the most “traditional” joint models, which generally consider a common variance and a time-invariant link.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Similar content being viewed by others

References

  • Baghfalaki T, Ganjali M (2015) A Bayesian approach for joint modeling of skew-normal longitudinal measurements and time to event data. REVSTAT-Stat J 13(2):169–191

    MATH  Google Scholar 

  • Brezger A, Lang S (2006) Generalized structured additive regression based on Bayesian p-splines. Comput Stat Data Anal 50:967–991

    MathSciNet  MATH  Google Scholar 

  • Carlin B, Louis T (2001) Bayes and empirical Bayes methods for data analysis, 2nd edn. Chapman & Hall, London

    MATH  Google Scholar 

  • Chen J, Huang Y (2015) A Bayesian mixture of semiparametric mixed-effects joint models for skewed-longitudinal and time-to-event data. Stat Med 34(20):2820–2843

    MathSciNet  Google Scholar 

  • Christensen R, Johnson W, Branscum A, Hanson TE (2011) Bayesian ideas and data analysis: an introduction for scientists and statisticians. Chapman & Hall/CRC Press, London

    MATH  Google Scholar 

  • Costa MJ, Shaw JEH (2009) Parametrization and penalties in spline models with an application to survival analysis. Comput Stat Data Anal 53(3):657–670

    MathSciNet  MATH  Google Scholar 

  • Cox D (1972) Regression models and life tables. J R Stat Soc B 34:187–220

    MathSciNet  MATH  Google Scholar 

  • Eilers PHC, Marx BD (2010) Splines, knots, and penalties. WIREs Comput Stat 2(6):637–653

    Google Scholar 

  • Farcomeni A, Viviani S (2014) Longitudinal quantile regression in the presence of informative dropout through longitudinal-survival joint modeling. Stat Med 34(7):1199–1213. https://doi.org/10.1002/sim.6393

    Article  MathSciNet  Google Scholar 

  • Faucett CL, Thomas DC (1996) Simultaneously modelling censored survival data and repeatedly measured covariates: a Gibbs sampling approach. Stat Med 15(15):1663–1685

    Google Scholar 

  • Gao F, Miller J, Xiong C, Beiser J, Gordon M (2011) A joint-modeling approach to assess the impact of biomarker variability on the risk of developing clinical outcome. Stat Method Appl 20:83–100

    MathSciNet  MATH  Google Scholar 

  • Gelman A (2006) Prior distributions for variance parameters in hierarchical models (comment on article by browne and draper). Bayesian Anal 1(3):515–534

    MathSciNet  MATH  Google Scholar 

  • Gelman A, Hwang J, Vehtari A (2014) Understanding predictive information criteria for Bayesian models. Stat Comput 24(6):997–1016

    MathSciNet  MATH  Google Scholar 

  • Guo X, Carlin P (2004) Separate and joint modelling of longitudinal and event time data using standard computer packages. Am Stat 58(1):16–24

    Google Scholar 

  • Hanson T, Branscum A, Johnson W (2011) Predictive comparison of joint longitudinal-survival modeling: a case study illustrating competing approaches. Lifetime Data Anal 17:3–28. https://doi.org/10.1007/s10985-010-9162-0

    Article  MathSciNet  MATH  Google Scholar 

  • Henderson R, Diggle P, Dobson A (2000) Joint modelling of longitudinal measurements and event time data. Biostatistics 1(4):465–480

    MATH  Google Scholar 

  • Hennerfeind A, Brezger A, Fahrmeir L (2006) Geoadditive survival models. JASA 101:1065–1075

    MathSciNet  MATH  Google Scholar 

  • Hofner B, Kneib T, Hartl W, Kuchenhoff H (2011) Building Cox-type structured hazard regression models with time-varying effects. Stat Model 11(1):3–24

    MathSciNet  MATH  Google Scholar 

  • Hsieh F, Tseng Y, Wang J (2006) Joint modeling of survival and longitudinal data: likelihood approach revisited. Biometrics 62:1037–1043

    MathSciNet  MATH  Google Scholar 

  • Huang Y, Hu XJ, Dagne GA (2014) Jointly modeling time-to-event and longitudinal data: a Bayesian approach. Stat Method Appl 23(1):95–121

    MathSciNet  MATH  Google Scholar 

  • Ibrahim J, Chen MH, Sinha D (2004) Bayesian methods for joint modeling of longitudinal and survival data with applications to cancer vaccine studies. Stat Sin 14(3):863–883

    MATH  Google Scholar 

  • Jiang B, Elliott MR, Sammel MD, Wang N (2015) Joint modeling of cross-sectional health outcomes and longitudinal predictors via mixtures of means and variances. Biometrics 71(2):487–97

    MathSciNet  MATH  Google Scholar 

  • Kalbfleisch JD, Prentice RL (2002) The statistical analysis of failure time data. John Wiley, New York

    MATH  Google Scholar 

  • Kneib T, Fahrmeir L (2007) A mixed model approach for geoadditive hazard regression. Scand J Stat 34(1):207–228

    MathSciNet  MATH  Google Scholar 

  • Lang S, Brezger A (2004) Bayesian p-splines. J Comput Graph Stat 13(1):183–212

    MathSciNet  MATH  Google Scholar 

  • Lange N, Carlin BP, Gelfand AE (1992) Hierarchical Bayes models for the progression of HIV infection using longitudinal CD4 T-cell numbers. J Am Stat Assoc 87(419):615–626

    MATH  Google Scholar 

  • Lin DY, Fischl MA, Schoenfeld DA (1993) Evaluating the role of CD4-lymphocyte counts as surrogate endpoints in human immunodeficiency virus clinical trials. Stat Med 12(9):835–842

    Google Scholar 

  • Lin X, Raz J, Harlow S (1997) Linear mixed models with heterogeneous within-cluster variances. Biometrics 53(3):910–923

    MathSciNet  MATH  Google Scholar 

  • Lunn D, Thomas A, Best N, Spiegelhalter D (2000) Winbugs—a bayesian modelling framework: concepts, structure, and extensibility. Stat Comput 10:325–337

    Google Scholar 

  • Lunn D, Jackson C, Best N, Thomas A, Spiegelhalter D (2013) The BUGS book: a practical introduction to Bayesian analysis. Chapman & Hall/CRC, London

    MATH  Google Scholar 

  • Lyles RH, Munõz A, Xu J, Taylor JMG, Chmiel JS (1999) Adjusting for measurement error to assess health effects of variability in biomarkers. Stat Med 18:1069–1086

    Google Scholar 

  • Martins R, Silva GL, Andreozzi V (2016) Bayesian joint modeling of longitudinal and spatial survival aids data. Stat Med 35(19):3368–3384. https://doi.org/10.1002/sim.6937

    Article  MathSciNet  Google Scholar 

  • Martins R, Silva G, Andreozzi V (2017) Joint analysis of longitudinal and survival aids data with a spatial fraction of long-term survivors: a Bayesian approach. Biom J 59(6):1166–1183

    MathSciNet  MATH  Google Scholar 

  • May MT, Gompels M, Delpech V, Porter K, Orkin C, Kegg S, Sabin C (2014) Impact on life expectancy of hiv-1 positive individuals of cd4+ cell count and viral load response to antiretroviral therapy. AIDS 28(8):1193–1202

    Google Scholar 

  • McLain AC, Lum KJ, Sundaram R (2012) A joint mixed effects dispersion model for menstrual cycle length and time-to-pregnancy. Biometrics 68(2):648–656

    MathSciNet  MATH  Google Scholar 

  • Rizopoulos D (2012) Joint models for longitudinal and time-to-event data with applications in R. Chapman and Hall/CRC, London

    MATH  Google Scholar 

  • Rizopoulos D (2016) The R package jmbayes for fitting joint models for longitudinal and time-to-event data using MCMC. J Stat Softw 72(7):1–46

    Google Scholar 

  • Rizopoulos D, Ghosh P (2011) A Bayesian semiparametric multivariate joint model for multiple longitudinal outcomes and a time-to-event. Stat Med 30(12):1366–1380

    MathSciNet  Google Scholar 

  • Rizopoulos D, Verbeke G, Lesaffre E (2009) Fully exponential Laplace approximations for the joint modelling of survival and longitudinal data. J R Stat Soc B 71:637–654

    MathSciNet  MATH  Google Scholar 

  • Rubin DB (1976) Inference and missing data. Biometrika 63(3):581–592

    MathSciNet  MATH  Google Scholar 

  • Song X, Wang C (2008) Semiparametric approaches for joint modeling of longitudinal and survival data with time-varying coefficients. Biometrics 64(2):557–566

    MathSciNet  MATH  Google Scholar 

  • Souza-Jr P, Szwarcwald CL, Castilho EA (2007) Delay in introducing antiretroviral therapy in patients infected by HIV in Brazil, 2003–2006. Clin Sci 62(5):579–584

    Google Scholar 

  • Spiegelhalter D, Best N, Carlin B, Van der Linde A (2002) Bayesian measures of model complexity and fit (with discussion). J R Stat Soc B 64(4):583–639

    MathSciNet  MATH  Google Scholar 

  • Tang NS, Tang AM, Pan DD (2014) Semiparametric Bayesian joint models of multivariate longitudinal and survival data. Comput Stat Data Anal 77:113–129

    MathSciNet  MATH  Google Scholar 

  • Tseng YK, Hsieh FS, Wang JL (2005) Joint modelling of accelerated failure time and longitudinal data. Biometrika 92(3):587–603

    MathSciNet  MATH  Google Scholar 

  • Watanabe S (2010) Asymptotic equivalence of Bayes cross validation and widely applicable information criterion in singular learning theory. J Mach Learn Res 11(455):3571–3591

    MathSciNet  MATH  Google Scholar 

  • Wulfsohn M, Tsiatis A (1997) A joint model for survival and longitudinal data measured with error. Biometrics 53:330–339

    MathSciNet  MATH  Google Scholar 

  • Yang M, Luo S, DeSantis S (2018) Bayesian quantile regression joint models: Inference and dynamic predictions. Stat Methods Med Res 28(8):2524–2537. https://doi.org/10.1177/0962280218784757

    Article  MathSciNet  Google Scholar 

  • Yu Z, Liu L, Bravata DM, Williams LS (2014) Joint model of recurrent events and a terminal event with time-varying coefficients. Biom J 56(2):183–197

    MathSciNet  MATH  Google Scholar 

  • Zhang H, Huang Y (2019) Quantile regression-based Bayesian joint modeling analysis of longitudinal-survival data, with application to an AIDS cohort study. Lifetime Data Anal 26(2):339–368. https://doi.org/10.1007/s10985-019-09478-w

    Article  MathSciNet  MATH  Google Scholar 

  • Zhu H, Ibrahim J, Chi YY, Tang N (2012) Bayesian influence measures for joint models for longitudinal and survival data. Biometrics 68(3):954–964

    MathSciNet  MATH  Google Scholar 

  • Zhuang R, Chen YQ (2019) Measuring surrogacy in clinical research. Stat Biosci pp 1–29

Download references

Acknowledgements

The author wish to thank Maria Goretti Fonseca and Cláudia Coeli for the database. This work was partially funded by Fundação para a Ciência e a Tecnologia projects UID/BIM/04585/2016 and UIDB/00006/2020.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Rui Martins.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Appendix

Appendix

Following Gelman et al. (2014) we define WAIC as

$$\begin{aligned} \text {WAIC}&= -2\left[ \text {lppd} - p_\text {WAIC}\right] \\&= -2\left[ \text {log pointwise predictive density} - \text { Penalization}\right] \\&= -2\left[ \log \left\{ \prod _{i=1}^N p(\mathcal {D}_i|\mathcal {D}) \right\} - \sum _{i=1}^N \mathbb {V}_{{\varvec{\theta }}|\mathcal {D}} \left[ \log \{p(\mathcal {D}_i|{\varvec{\theta }})\} \right] \right] \\&= -2\left[ \sum _{i=1}^N \log \left\{ \int p(\mathcal {D}_i|{\varvec{\theta }})\pi ({\varvec{\theta }}|\mathcal {D}) \right\} d{\varvec{\theta }} \, - \sum _{i=1}^N V_{s=1}^S \left[ \log \{ p(\mathcal {D}_i|{\varvec{\theta }}^s) \} \right] \right] \\&\approx -2 \Biggl [ \sum _{i=1}^N \log \left\{ \frac{1}{S} \sum _{s=1}^S p(\mathcal {D}_i | {\varvec{\theta }}^s) \right\} - \sum _{i=1}^N \frac{1}{S-1} \sum _{s=1}^S \Biggl ( \log \{p(\mathcal {D}_i | {\varvec{\theta }}^s)\} - \\&\quad \left\{ \frac{1}{S} \sum _{s=1}^S \log p(\mathcal {D}_i | {\varvec{\theta }}^s) \right\} \Biggr )^{ 2} \Biggr ] \\&= -2\left[ \sum _{i=1}^N \log \left\{ \overline{p(\mathcal {D}_i | {\varvec{\theta }})} \right\} - \sum _{i=1}^N \frac{1}{S-1} \sum _{s=1}^S \left( \log \{p(\mathcal {D}_i | {\varvec{\theta }}^s)\} - \overline{\log \{p(\mathcal {D}_i | {\varvec{\theta }}) \} } \right) ^2 \right] \end{aligned}$$

where \(V_{s=1}^S [a_s]\), represents the posterior sample variance and the over line stands for the posterior sample mean. Like the DIC models with smaller WAIC values should be preferred.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Martins, R. A flexible link for joint modelling longitudinal and survival data accounting for individual longitudinal heterogeneity. Stat Methods Appl 31, 41–61 (2022). https://doi.org/10.1007/s10260-021-00566-6

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10260-021-00566-6

Keywords

Navigation