In many biomedical studies, repeated measures are frequently available (e.g., daily stress, blood sugar measurements, caffeine consumption, daily hormonal profile) and linked to health outcomes (e.g., time to pregnancy, developmental milestone, labor arrests). However, several features of these longitudinal data should be considered before making such inferences. The focus of this area of research is to build models for the longitudinal process based on informed underlying biological/clinical knowledge and use joint modeling of repeatedly measured and time-to-event data to help assess dynamics of the event progression and to derive personalized prognosis.
Joint modeling combines longitudinal models and time-to-event models under different dependence structure to relate subject-specific trajectory to their prognosis. The goal is to describe several aspects of the relationship between time-varying longitudinal markers and the endpoint of interest. This provides valuable statistical tools for study purposes as well as helps healthcare providers in making well-informed dynamic medical decisions. Multiple studies on human fertility as well as labor progression and adolescent driving studies provide motivation for this area of research.
Rajeshwari Sundaram, M.Stat., Ph.D.
Lum, K. J., Sundaram, R., Buck, L. G. M., & Louis, T. A. (2016). A Bayesian joint model of menstrual cycle length and fecundity. Biometrics, 72(1):193-203. PMID: 26295923. PMCID: PMC4761533
McLain, A. C., Sundaram, R., & Buck, L. G. M. (2015). Joint analysis of longitudinal and survival data measured on nested timescales using shared parameter models: an application to fecundity data. Journal of the Royal Statistical Society: Series C (Applied Statistics), 64(2), 339-357. PMID: 27122641. PMCID: PMC4844229
Katki, H. A., Cheung, L. C., Fetterman, B., Castle, P. E., & Sundaram, R. (2015). A joint model of persistent human papillomavirus infection and cervical cancer risk: implications for cervical cancer guidelines. Journal of the Royal Statistical Society: Series A (Statistics in Society), 178(4), 903-923. PMID: 26556961. PMCID: PMC4635446
McLain, A. C., Lum, K. J., & Sundaram, R. (2012). A joint mixed effects dispersion model for menstrual cycle length and time-to-pregnancy. Biometrics, 68(2), 648-656. PMID: 22321128
Kim, S., Sundaram, R., & Buck, L. G. M. (2010). Joint modeling of intercourse behavior and human fecundability using structural equation models. Biostatistics, 11(3), 559-571. PMID: 20173100. PMCID: PMC2912701