A 12-week structured self-management program, followed by a 12-week consolidation and remote contacts, produced measurable clinical changes among patients with type 2 diabetes identified as high risk for Diabetic Kidney Disease. Investigators reported statistically significant between-group improvements in HbA1c, eGFR, systolic blood pressure and LDL-C compared with usual care. The trial team observed no significant changes for urinary albumin-to-creatinine ratio, BMI or waist circumference.
The study was a prospective, two-arm, assessor-blinded randomized controlled trial conducted at a tertiary hospital in Shandong Province, China. The trial enrolled 134 eligible adults with type 2 diabetes who scored at or above the predefined DKD risk threshold using a validated prediction model. Recruitment ran from 2 December 2024 through 28 February 2025. Randomization used pre-generated numbers placed in sealed envelopes; odd numbers allocated to the intervention arm and even numbers to control. Outcome assessors and participants were kept unaware of assignments, while blinding of intervention staff and data analysts was not feasible.
The intervention team comprised an endocrinologist, a nursing manager, two diabetes-specialized nurses and a nursing researcher. The self-management program covered five core topics: DKD knowledge, diet management, physical activity, medication management and self-monitoring. Content was scripted from authoritative guidelines and framed using the Health Belief Model to promote perceived risk recognition and strengthen self-efficacy.
Delivery used six video-based sessions over a six-week initial learning phase, followed by a strengthening phase. After completing the program, participants received monthly telephone or WeChat contacts during a 12-week follow-up to monitor adherence and support behavior maintenance. The control group received usual inpatient-oriented health education and routine medical care as delivered by physicians and primary nurses.
Clinical indicators were abstracted from medical records at three prespecified time points: baseline (T0), post-intervention at 12 weeks (T1) and at the end of follow-up at 24 weeks (T2). Outcomes included HbA1c, urinary albumin-to-creatinine ratio (UACR), estimated glomerular filtration rate (eGFR), systolic blood pressure (SBP), low-density lipoprotein cholesterol (LDL-C), body mass index (BMI) and waist circumference.
Analysts applied generalized estimating equations to model repeated measures and to test group-by-time interaction effects, specifying an exchangeable correlation structure and using robust standard errors. Analyses followed the intention-to-treat principle. For skewed variables, log-transformation was performed before modeling and back-transformed for interpretation. To evaluate robustness to missingness, the team conducted multiple imputation (10 datasets) and pooled results using standard combining rules; sensitivity analyses were reported as consistent with primary findings.
From 317 screened patients, 134 met eligibility and consented to enrollment. All participants completed the active intervention phase. During follow-up, attrition occurred: 7 intervention and 9 control participants were lost by T1, and 5 intervention and 4 control participants were lost by T2. The investigators tested for baseline differences between retained participants and those lost to follow-up and found no statistically significant differences (p > 0.05), which the team interpreted as consistent with data missing at random.
Compared with usual care, the intervention group showed significant improvements in glycemic control and several cardiometabolic and renal markers. Investigators reported reductions in HbA1c, improvements in eGFR, lower SBP and decreased LDL-C over the 24-week assessment window. Sensitivity analyses using multiple imputation produced similar effect estimates, reinforcing the primary analytic conclusions.
No significant between-group differences were observed for urinary albumin-to-creatinine ratio (UACR), body mass index or waist circumference during the 24-week study period. The study team did not report additional subgroup or mechanistic analyses beyond the primary repeated-measures models in the source article.
The investigators documented several study limitations in the report. Recruitment occurred at a single tertiary center, which may limit external generalizability. Follow-up extended only to 24 weeks; the durability of observed changes beyond that interval was not reported. Blinding of those delivering the intervention and of data analysts was not feasible, introducing potential performance and analytic bias despite assessor blinding. The source did not report longer-term renal endpoints such as progression to microalbuminuria conversion rates beyond 24 weeks, nor did it provide detailed per-protocol adherence metrics in the public article text. These unanswered questions — the long-term trajectory of UACR and whether short-term improvements translate into reduced DKD incidence — remain to be addressed in subsequent work.