Research Article
Socio-economic, Demographic, and Clinical Predictors of Diabetic Kidney Disease Progression (Renal Function Decline) Among Adults with Diabetes: A Retrospective Cohort Study in Kenya
Issue:
Volume 15, Issue 2, April 2026
Pages:
27-39
Received:
1 February 2026
Accepted:
20 February 2026
Published:
5 March 2026
Abstract: Diabetic kidney disease (DKD) represents a major global health burden, yet predictive models often overlook socio-economic determinants that may independently influence disease progression. This retrospective cohort study aimed to identify socio-economic, demographic, and clinical predictors of diabetic kidney disease development among diabetic patients in Kenya and compare Cox regression with support vector machine (SVM) models for risk prediction. Data were collected from 756 adult diabetic patients attending Meru Teaching and Referral Hospital and Kerugoya Level 5 Hospital between January 2018 and July 2024 through medical record review and semi-structured questionnaires. Survival analysis employed Kaplan-Meier estimation, log-rank tests, multivariable Cox proportional hazards regression, and survival SVM modeling. During follow-up, 286 participants (37.8%) developed diabetic kidney disease. Multivariable Cox analysis identified seven significant predictors of diabetic kidney disease progression: older age at diabetes diagnosis (adjusted HR=1.023, p=0.002), male gender (HR=1.282, p=0.041), family history of chronic kidney disease (HR=6.919, p<0.001), alcohol consumption (HR=1.556, p=0.001), and financial hardship (HR=4.524, p<0.001) increased risk, while secondary/higher education (HR=0.593, p<0.001) and ever being employed (HR=0.635, p=0.011) were protective. The SVM model demonstrated marginally superior predictive accuracy (C-index=0.775) versus Cox regression (C-index=0.770). These findings underscore that socio-economic factors function as independent risk modifiers beyond traditional clinical parameters, challenging conventional prediction paradigms that focus exclusively on biomedical indicators. The high incidence of diabetic kidney disease observed highlights an urgent public health challenge requiring integrated screening protocols that assess both clinical and socio-economic risk profiles at diabetes diagnosis. We recommend implementing targeted public health interventions that address financial barriers, promote educational attainment, and support employment opportunities for diabetic patients to mitigate diabetic kidney disease progression in resource-limited settings.
Abstract: Diabetic kidney disease (DKD) represents a major global health burden, yet predictive models often overlook socio-economic determinants that may independently influence disease progression. This retrospective cohort study aimed to identify socio-economic, demographic, and clinical predictors of diabetic kidney disease development among diabetic pat...
Show More
,
Moses Muraya
