Incidence and predictors of chronic kidney disease among hypertensive patients in Ethiopia: A Bayesian multivariate joint model.

Background: Hypertension (HTN) is a major contributor to chronic kidney disease (CKD), a leading cause of global morbidity and mortality. In low-resource settings such as Ethiopia, where CKD risk factors remain understudied, identifying predictors and longitudinal blood pressure (BP) patterns associated with CKD incidence is crucial for early intervention. Therefore, this study aimed to determine the incidence and predictors of CKD, as well as its association with longitudinal BP changes, among hypertensive patients in Ethiopia.

Methods: An institution-based retrospective follow-up study was conducted at the University of Gondar Comprehensive Specialized Referral Hospital. Using a Bayesian joint modeling approach, we integrated Cox proportional hazard and linear mixed effects models to evaluate the effects of time-dependent BP trajectories on CKD risk. The data were entered into the Kobo toolbox and analyzed with R software (version 4.3.1).

Results: A total of 408 hypertensive patients were followed for 2322.83 person-years. At the end of the follow-up, 58/408 (14.22%) developed CKD, with an incidence density of 2.5 cases per 100 person-years (95% CI: 1.89-3.14). Both the current values and longitudinal the quarterly rate of change in BP were significantly associated with increased CKD risk. For systolic BP, the adjusted hazard ratio (AHR) was 6.25 (95% CrI: 2.85-9.85) for the current values and 3.75 (95% CrI: 3.16-7.95) for the quarterly rate of change. Similarly, the diastolic BP had an AHR of 4.32 (95% CrI: 2.35-8.27) for the current values and 5.64 (95% CrI: 4.24-10.82) for the quarterly rate of change. Additionally, age ≥ 65 years (AHR = 4.62; 95% CrI: 1.83-12.21), HDL-C < 40 mg/dL (AHR = 3.32; 95% CrI: 1.73-7.86), diabetes mellitus (AHR = 3.08; 95% CrI: 2.01-9.54), and proteinuria positivity (AHR = 2.85; 95% CrI: 1.48-5.55) were significant predictors of the incidence of CKD. These findings highlight the importance of close BP monitoring in Ethiopian hypertension clinics.

Conclusion: The incidence of CKD in this study was relatively high compared with that reported in previous similar studies conducted in Ethiopia. Our findings confirm that time-dependent systolic BP and diastolic BP trajectories are strongly associated with an increased risk of CKD. Additionally, age, low HDL-C levels (<40 mg/dl), the presence of diabetes mellitus, and proteinuria were identified as significant predictors of CKD. Therefore, effective CKD prevention among hypertensive patients in Ethiopia hinges on regularly checking both their current blood pressure levels and how those levels change over time. We also need to keep a close eye on older patients (65 + years), low HDL-C, diabetes, and proteinuria to catch those at highest risk early and step in with care.