Abstract:
Background:- Diabetes mellitus is currently a public health issue in both developed and
developing nations and is linked to greater morbidity and mortality globally. We aim to model
the progression of several biomarkers, including SBP, FBS, and DBP, as well as the correlation
between various aspects of this progression and the time to develop microvascular complications
in various association structures. Methods: A retrospective follow-up study was conducted on co-infected patients at Tepi
General Hospital and MTUTH. Ethiopia with an enrollment time of September 2015 and
followed for four years. We presented a joint model that consists of two related sub-models: a
multivariate linear mixed model for FBS, SBP, and DBP level and a survival sub-model for the
patient's developing retinopathy status. Three different forms of joint modeling in the Bayesian
approach via R-statistical software were used. For the analysis, we used the JMBayes Markov
Chain Monte Carlo (MCMC) algorithm to sample from the posterior conditional distributions. Results: Time had a significant effect on all repeated measurements (SBP, FBS, and DBP)
(P<0.001) in the multivariate linear mixed model. The square root FBS, SBP, and DBP are
statistically significant at a 5% level of significance in all forms of the joint models that we
considered: This indicates that there is strong evidence of the association between the effect of
the longitudinal square root FBS, SBP and DBP on the risk of developing retinopathy for Co infected patients. Time-updated FBS of Co Infected patients makes them develop complications
earlier. Conclusion: At a 5% level of significance(p-value<0.05), the risk of patients' hazard become
increases as the SBP, FBS, and DBP become increase. From a methodological perspective
considering different association structures in joint modeling leads to appropriate decisions. The
fixed effects of baseline age, residence, educational status, occupation, smoking, family history
of DM, and weight significantly affect the developing retinopathy time(p-value<0.05).
