Type 2 diabetes has become a pandemic disorder and its increase in prevalence raises concerns worldwide [
The association of microalbuminuria and DR has been reported in patients with type 1 diabetes. Among patients with type 1 diabetes who have nephropathy, more than 95% already have DR [
Several studies have shown that DR severity was significantly associated with reduced kidney function and increased risk of CKD in type 2 diabetes [
We enrolled patients with type 2 diabetes from the diabetes clinic in the Department of Endocrinology of Kangnam Sacred Heart Hospital who underwent fundus photographic examinations for DR and whose renal profiles were studied between August 2006 and February 2014. We excluded patients with
Baseline characteristics, including demographics (age, gender), medical history (hypertension, diabetes, duration of diabetes), and laboratory variables were collected at the time of first DR assessment. Weight and height were assessed and body mass index (BMI) was calculated. Blood pressure was measured with a sphygmomanometer after 5 min of rest. Hemoglobin A1c (HbA1c) was measured using a method that was NGSP certified and standardized to the DCCT assay. A standard urine dipstick was used to measure albuminuria qualitatively. Serum creatinine was measured using the modified Jaffe method. Based on the serum creatinine concentration, the eGFR was calculated using the four-variable equation from the Modification of Diet in Renal Disease study [
DR presence was assessed by retinal ophthalmologists who had no knowledge of the clinical details using slit-lamp examination, indirect ophthalmoscopy, and/or fluorescein angiography. Patients were classified into the following categories: (1) normal: no apparent sign of DR; (2) non-proliferative DR (NPDR): including microaneurysms, hard exudates, intraretinal hemorrhages, venous beading, or prominent intraretinal microvascular abnormality; and (3) proliferative DR (PDR): including retinal or optic disk neovascularization, vitreous hemorrhage, or preretinal hemorrhage, according to the Global Diabetic Retinopathy Project Group [
Data were expressed as mean (standard deviation) for continuous variables and as numbers of cases and percentages for categorical variables. Patients were stratified by DR grade and progression of DR status. Differences between the groups were assessed using the chi-squared test for dichotomous factors and one-way ANOVA for continuous factors. Logistic regression analyses with stepwise variable selection were performed to assess the independent association of progression of DR. Univariate logistic regression models were employed first, followed by multivariate logistic regression models with adjustment by covariates that were significant (
Clinical characteristics of the patients according to the retinopathy group are shown in Table
Baseline characteristics of type 2 diabetes according to retinopathy status.
Characteristics | All participants, | No DR, | NPDR, | PDR, | |
---|---|---|---|---|---|
Age (years) | 0.003 | ||||
Male | 724 (47.4) | 418 (46.5) | 188 (46.4) | 118 (52.7) | 0.232 |
Duration of diabetes (years) | <0.001 | ||||
BMI (kg/m2) | <0.001 | ||||
SBP (mmHg) | 0.014 | ||||
DBP (mmHg) | 0.063 | ||||
FPG (mg/dl) | <0.001 | ||||
HbA1c (%) | <0.001 | ||||
Hemoglobin (g/dl) | <0.001 | ||||
Total cholesterol (mg/dl) | 0.826 |
Data are expressed as
Figure
Prevalent diabetic retinopathy and renal function. (a) eGFR category: G1 (≥90), G2 (60–89), G3 (30–59), and G4 (15–29 ml/min/1.73m2) and (b) albuminuria. Abbreviation: eGFR: estimated glomerular filtration rate.
The mean follow-up period was
Comparison of characteristics by progression of diabetic retinopathy (
Characteristics | Non-progression, | Progression, | |
---|---|---|---|
Age (years) | 0.060 | ||
Male | 545 (46.6) | 61 (45.5) | 0.809 |
Duration of diabetes (years) | 10.6 ± 7.8 | 0.002 | |
BMI (kg/m2) | 0.560 | ||
SBP (mmHg) | 0.467 | ||
DBP (mmHg) | 0.433 | ||
FPG (mg/dl) | <0.001 | ||
HbA1c (%) | <0.001 | ||
eGFR (ml/min/1.73m2) | 0.520 | ||
eGFR category | |||
G1 (≥90 ml/min/1.73m2) | 303 (25.9) | 34 (25.4) | 0.483 |
G2 (60-89 ml/min/1.73m2) | 664 (56.8) | 70 (52.2) | |
G3 (30-59 ml/min/1.73m2) | 176 (15.1) | 27 (20.1) | |
G4 (15-29 ml/min/1.73m2) | 26 (2.2) | 3 (2.2) | |
Albuminuria | |||
- | 979 (91.8) | 95 (81.2) | <0.001 |
+~+++ | 88 (8.2) | 22 (18.8) | |
Hemoglobin (g/dl) | 0.782 | ||
Total cholesterol (mg/dl) | 0.581 |
Data are expressed as
Figure
Changes of renal function in both groups. (a) Mean eGFR change. (b) Rate of ≥20% decline in eGFR during the follow-up period. (c) Albuminuria development. Abbreviation: eGFR: estimated glomerular filtration rate.
We performed binary logistic regression analysis to identify factors associated with progression of DR. On univariate analysis, longer duration of diabetes (odds ratio (OR) 1.034, 95% CI 1.012–1.057,
Logistic models for DR progression.
Characteristics | From no DR to NPDR | From NPDR to PDR | ||
---|---|---|---|---|
Univariate | Multivariate | Univariate | Multivariate | |
Odds ratio (95% CI) | Odds ratio (95% CI) | Odds ratio (95% CI) | Odds ratio (95% CI) | |
Age (per year) | 0.999 (0.981-1.018) | 0.953 (0.926-0.980) | 0.966 (0.937-0.995) | |
Duration of diabetes (per year) | 1.067 (1.037-1.098) | 1.062 (1.025-1.100) | 0.995(0.995-1.037) | |
FPG (per mg/dl) | 1.006 (1.003-1.008) | 1.003 (1.000-1.007) | 1.003 (0.999-1.007) | |
HbA1c (per %) | 1.401 (1.261-1.557) | 1.353 (1.191-1.537) | 1.210 (1.028-1.426) | 1.111 (0.933-1.323) |
Albuminuria (vs. no) | 3.843 (1.889-7.819) | 2.791 (1.244-6.263) | 1.993 (0.908-4.376) | |
eGFR (per ml/min/1.73m2) | 0.993 (0.982-1.003) | 0.998 (0.983-1.013) | ||
Decrease | 1.953 (1.005-3.796) | 0.879 (0.373-2.075) | 3.423 (1.696-6.908) | 2.553 (1.219-5.348) |
This study evaluated the association of progression of DR and declining renal function in patients with type 2 diabetes. We found that prevalent DR severity was associated with decreased eGFR and albuminuria. Furthermore, a ≥20% decline in eGFR was independently associated with the progression of NPDR to PDR. Duration of diabetes, albuminuria, and HbA1c were independent risk factors for progression of no DR to NPDR. The results of this study suggest that investigation of DR status should be recommended for patients with declining renal function, especially for NPDR patients. This result also supports the notion of a shared pathogenetic mechanism of DR and diabetic nephropathy in patients with type 2 diabetes.
Both the retina and the kidney are supplied by very small vessels. The anatomical similarities in the vascularization of the retina and the kidney give rise to complications of diabetes in the small vessels that appear in both organs. The microvascular changes in both organs are thought to be initiated by chronic hyperglycemia, followed by the progressive narrowing and eventual occlusion of vascular lumina, subsequently leading to inadequate perfusion of affected tissues [
There are several studies about the association between DR and kidney function [
We showed that risk factors of DR development were different from ones of DR progression. The difference can be explained to originate from difference of their natural course and clinical characteristics. There are few studies to explore the DR natural course. In the previous studies, development from no DR to NPDR was estimated to take about 14 years, while DR progression rate was considerably fast, around 4 years, for patients who progressed to sever e form of DR [
Few cohort studies have investigated progression of DR in patients with type 2 diabetes, and most of these found that microalbuminuria, duration of diabetes, glycemic level, and baseline blood pressure were important independent predictors of DR incidence and progression [
This study had several limitations. First, there is the inherent weakness of all studies with a retrospective design, namely, the use of data from past medical records. Thus, we cannot propose causal associations or prediction of declining renal function in patients with DR. Future prospective cohort studies are needed to determine whether the progression of renal disease predicts progression of DR in patients with type 2 diabetes. Second, DR classification in this study was based on graders’ discretion which might result in potential bias. However, each experienced retinal specialist determined DR grades according to globally accepted guideline. Third, we cannot know for certain whether risk factors are causing retinopathy progression or merely represent markers of disease progression. Fourth, because we were not able to collect dipstick results from all participants, we performed multivariate analysis for progression of DR excluding those without dipstick results. Furthermore, we could not obtain quantitative results of albuminuria such as urine microalbumin-to-creatinine ratio. Fifth, we did not collect CKD duration data and could not analyze its contribution as a prognostic factor.
In conclusion, we identified several factors associated with the progression of DR, including HbA1c level, duration of diabetes, younger age, and declining renal function. These factors were associated with DR progression in different ways, depending on baseline DR status. The decrease of renal function was associated with progression of DR, especially in patients with NPDR. This result supports the notion that an individualized screening schedule according to the individual patient’s risk might be needed. Future prospective cohort studies are needed to evaluate the predictive value of renal disease for the development and progression of DR.
The authors confirm that the data supporting the findings of this study are available within its supplementary materials.
The authors report no conflict of interest.
AJ.C. wrote the manuscript and researched data. H.P. researched data and contributed to the discussion. Y.L. reviewed/edited the manuscript and researched data. Y.S. researched data and contributed to the discussion. S.B. researched data. H.K. researched data. AJ.C. and Y. L are included in the writing group.