Clinical features associated with a rapid decline in renal function among Japanese patients with type 2 diabetes mellitus: microscopic hematuria coexisting with diabetic retinopathy

Prediction of the risk of developing end-stage renal diseases in newly diagnosed type 2 diabetes mellitus using artificial intelligence algorithms

OBJECTIVES

Type 2 diabetes mellitus (T2DM) imposes a great burden on healthcare systems, and these patients experience higher long-term risks for developing end-stage renal disease (ESRD). Managing diabetic nephropathy becomes more challenging when kidney function starts declining. Therefore, developing predictive models for the risk of developing ESRD in newly diagnosed T2DM patients may be helpful in clinical settings.

METHODS

We established machine learning models constructed from a subset of clinical features collected from 53,477 newly diagnosed T2DM patients from January 2008 to December 2018 and then selected the best model. The cohort was divided, with 70% and 30% of patients randomly assigned to the training and testing sets, respectively.

RESULTS

The discriminative ability of our machine learning models, including logistic regression, extra tree classifier, random forest, gradient boosting decision tree (GBDT), extreme gradient boosting (XGBoost), and light gradient boosting machine were evaluated across the cohort. XGBoost yielded the highest area under the receiver operating characteristic curve (AUC) of 0.953, followed by extra tree and GBDT, with AUC values of 0.952 and 0.938 on the testing dataset. The SHapley Additive explanation summary plot in the XGBoost model illustrated that the top five important features included baseline serum creatinine, mean serum creatine within 1 year before the diagnosis of T2DM, high-sensitivity C-reactive protein, spot urine protein-to-creatinine ratio and female gender.

CONCLUSIONS

Because our machine learning prediction models were based on routinely collected clinical features, they can be used as risk assessment tools for developing ESRD. By identifying high-risk patients, intervention strategies may be provided at an early stage.

Hematuria as a risk factor for progression of chronic kidney disease and death: findings from the Chronic Renal Insufficiency Cohort (CRIC) Study

BACKGROUND

Hematuria is associated with chronic kidney disease (CKD), but has rarely been examined as a risk factor for CKD progression. We explored whether individuals with hematuria had worse outcomes compared to those without hematuria in the CRIC Study.

METHODS

Participants were a racially and ethnically diverse group of adults (21 to 74 years), with moderate CKD. Presence of hematuria (positive dipstick) from a single urine sample was the primary predictor. Outcomes included a 50% or greater reduction in eGFR from baseline, ESRD, and death, over a median follow-up of 7.3 years, analyzed using Cox Proportional Hazards models. Net reclassification indices (NRI) and C statistics were calculated to evaluate their predictive performance.

RESULTS

Hematuria was observed in 1145 (29%) of a total of 3272 participants at baseline. Individuals with hematuria were more likely to be Hispanic (22% vs. 9.5%, respectively), have diabetes (56% vs. 48%), lower mean eGFR (40.2 vs. 45.3 ml/min/1.73 m2), and higher levels of urinary albumin > 1.0 g/day (36% vs. 10%). In multivariable-adjusted analysis, individuals with hematuria had a greater risk for all outcomes during the first 2 years of follow-up: Halving of eGFR or ESRD (HR Year 1: 1.68, Year 2: 1.36), ESRD (Year 1: 1.71, Year 2: 1.39) and death (Year 1:1.92, Year 2: 1.77), and these associations were attenuated, thereafter. Based on NRIs and C-statistics, no clear improvement in the ability to improve prediction of study outcomes was observed when hematuria was included in multivariable models.

CONCLUSION

In a large adult cohort with CKD, hematuria was associated with a significantly higher risk of CKD progression and death in the first 2 years of follow-up but did not improve risk prediction.

Association between dipstick hematuria and decline in estimated glomerular filtration rate among Japanese patients with type 2 diabetes: A prospective cohort study [Diabetes Distress and Care Registry at Tenri (DDCRT 14)]

AIMS

To assess the association between dipstick hematuria and estimated glomerular filtration rate (eGFR) decline in Japanese patients with type 2 diabetes.

METHODS

Longitudinal data were obtained from 3068 Japanese patients with type 2 diabetes. To assess the independent association between dipstick hematuria and eGFR decline, we used Cox proportional hazard model adjusted for potential confounders.

RESULTS

Median follow-up period was 699.7days. Mean age, body mass index (BMI), and HbA1c level were 65.7years, 24.6kg/m², and 7.5% (58.1mmol/mol), respectively. Positive dipstick hematuria was significantly associated with baseline eGFR and severity of albuminuria (p<0.001). The multivariable-adjusted hazard ratio for eGFR decline in patients with dipstick hematuria compared with those without dipstick hematuria was 2.19 [95% confidence interval (CI): 1.22-3.91]; this association remained significant even after the exclusion of patients who did not have diabetic retinopathy (hazard ratio: 2.39; 95% CI: 1.13-5.04).

CONCLUSION

Positive dipstick hematuria was associated with severity of albuminuria and renal function. A significant association was found between dipstick hematuria and increased risk of eGFR decline among patients with type 2 diabetes. Therefore, our results suggest that dipstick hematuria is perhaps indicative of more severe diabetic nephropathy.

Risks of rapid decline renal function in patients with type 2 diabetes

Progressive rising population of diabetes and related nephropathy, namely, diabetic kidney disease and associated end stage renal disease has become a major global public health issue. Results of observational studies indicate that most diabetic kidney disease progresses over decades; however, certain diabetes patients display a rapid decline in renal function, which may lead to renal failure within months. Although the definition of rapid renal function decline remained speculative, in general, it is defined by the decrease of estimated glomerular filtration rate (eGFR) in absolute rate of loss or percent change. Based on the Kidney Disease: Improving Global Outcomes 2012 clinical practice guidelines, a rapid decline in renal function is defined as a sustained decline in eGFR of > 5 mL/min per 1.73 m(2) per year. It has been reported that potential factors contributing to a rapid decline in renal function include ethnic/genetic and demographic causes, smoking habits, increased glycated hemoglobin levels, obesity, albuminuria, anemia, low serum magnesium levels, high serum phosphate levels, vitamin D deficiency, elevated systolic blood pressure, pulse pressure, brachial-ankle pulse wave velocity values, retinopathy, and cardiac autonomic neuropathy. This article reviews current literatures in this area and provides insight on the early detection of diabetic subjects who are at risk of a rapid decline in renal function in order to develop a more aggressive approach to renal and cardiovascular protection.