Factors associated with progression of diabetic nephropathy in Japanese elderly patients with type 2 diabetes: Sub-analysis of the Japanese Elderly Diabetes Intervention Trial

Abstract

Aim:  Diabetic nephropathy is a serious complication in patients with type 2 diabetes. The aim of this study was to explore the factors associated with the progression of this complication in elderly patients with type 2 diabetes.

Methods:  This retrospective study of a subgroup of patients registered with the Japanese Elderly Diabetes Intervention Trial included 621 Japanese patients with type 2 diabetes mellitus (age ≥65 years, 346 with normoalbuminuria, 190 with microalbuminuria and 85 with overt proteinuria). Multivariate Cox proportional hazard regression model with a backward stepwise procedure was applied to select factors with significant effects on worsening of nephropathy stage and the doubling of serum creatinine.

Results:  During the follow up (median 52 months), 21% of patients progressed from normoalbuminuria and microalbuminuria to a worse nephropathy stage. Aging, female sex and high-density lipoprotein cholesterol were identified as independent and significant factors that worsen nephropathy stage. Also, 6.1% of patients showed doubling of serum creatinine during follow up. A positive history of cardiovascular disease, hyperuricemia and conventional therapy were identified as significant factors involved in the doubling of serum creatinine. The cumulative incidence of the doubling of serum creatinine was significantly lower in the intensive therapy group than the conventional therapy group (P = 0.016), although that of progression of nephropathy stage was similar in the two groups.

Conclusions:  We identified several factors associated with the progression of diabetic nephropathy in elderly patients with type 2 diabetes. The results suggest that multiple risk factor intervention seems important in preventing deterioration of renal dysfunction. Geriatr Gerontol Int 2012; 12 (Suppl. 1): 127–133.

Introduction

The number of patients who require chronic hemodialysis is progressively increasing in Japan. The annual report of the Japanese Society for Dialysis Therapy described a trend of a gradual increase in the mean age of new patients requiring dialysis, which was 67.2 years at the end of 2008,¹ reflecting the aging society in Japan. Diabetic nephropathy is a leading cause of end-stage kidney disease (ESKD). Type 2 diabetes is an age-related disorder with prevalence estimated at more than 5% of the population in Japan.² In addition, the majority of elderly patients with diabetic nephropathy also have various vascular diseases, further worsening the overall prognosis. Thus, halting the progression of diabetic nephropathy in elderly patients with type 2 diabetes is important clinically to improve prognosis. However, there is little information or evidence of this complication in elderly patients with type 2 diabetes. Understanding the clinical characteristics and factors associated with the progression of diabetic nephropathy in this population should be useful for establishing effective therapeutic strategies to prevent this life-threatening complication.

Elevated urinary albumin excretion, or microalbuminuria, is an early sign of diabetic nephropathy.³ During the natural course of this complication, diabetic patients with microalbuminuria progress to overt proteinuria and then to clinically-evident renal dysfunction.³ Several large-scale clinical trials have been carried out to prevent the progression of diabetic nephropathy. However, only a few such trials included elderly patients with type 2 diabetes.

The Japanese Elderly Diabetes Intervention Trial (J-EDIT), a nationwide, randomized, controlled, prospective, intervention trial for Japanese elderly patients with type 2 diabetes, was designed to clarify whether therapeutic intervention modulates the incidence of diabetic macro- and microvasculopathies in this population of type 2 diabetes.^4–6 Using the data of this J-EDIT, we carried out a subanalysis to define the clinical characteristics and factors associated with the progression of diabetic nephropathy as determined by increase in albuminuria and doubling of serum creatinine. We also investigated the effect of the multiple risk factor intervention on the progression of diabetic nephropathy.

Methods

Participants

Patients in the subanalysis were recruited from among the participants of the J-EDIT, which includes a registry of 1173 Japanese elderly patients with type 2 diabetes mellitus from 39 institutions and hospitals across Japan.^4–6 Details of the study design, protocol, randomization and intervention procedures of the J-EDIT were described by Araki et al.⁶ Briefly, the J-EDIT included only Japanese patients with type 2 diabetes who were 65 years or older at study entry and whose glycated hemoglobin A1c (HbA1c) levels were ≥7.4%. The values of HbA1c were expressed in National Glycohemoglobin Standardization Program values (NGSP), according to the recommendations of the Japanese Diabetes Society (JDS).⁷ The HbA1c (%) is estimated as a NGSP equivalent value calculated by the formula HbA1c = HbA1c (JDS) + 0.4%. Participants in the J-EDIT were randomly assigned to either the intensive therapy arm or the conventional therapy arm and were followed up from 2001 to 2007 or until the incidence of cardiovascular events. The treatment goal in the intensive therapy arm was HbA1c < 6.9%, body mass index (BMI) < 25 kg/m², blood pressure (BP) < 130/85 mmHg, high-density lipoprotein (HDL) cholesterol > 40 mg/dL, triglycerides < 150 mg/dL and total cholesterol < 180 mg/dL (or low-density lipoprotein [LDL] cholesterol < 100 mg/dL) if patients had any histories of coronary heart disease, or total cholesterol <200 mg/dL (or LDL cholesterol < 120 mg/dL) if not. Patients with a history of cerebral infarction were prescribed antiplatelet medications. Patients assigned to the conventional therapy arm continued the standard treatment of diabetes, hypertension and dyslipidemia without aiming at specific treatment goal. At study entry (baseline), the urinary albumin-creatinine ratio (UACR) in a spot urine sample was measured. According to the level of UACR, the stages of diabetic nephropathy were classified as normoalbuminuria (UACR: <30 mg/g creatinine), microalbuminuria (UACR 30–299 mg/g creatinine) and overt proteinuria (UACR: ≥300 mg/g creatinine). To ensure the reliable stage of diabetic nephropathy at baseline, we enrolled only patients whose stage of nephropathy was identical in two assessments at baseline. Of the total 1173 elderly patients, 340 were excluded as a result of inconsistencies in two stages of diabetic nephropathy assessed at baseline. Furthermore, the baseline data on UACR were not available in 143 patients, and thus their cases were excluded. In addition, we also excluded 69 patients whose follow-up data were incomplete or did not experience cardiovascular events before the end of the first year of follow up. Thus, a total of 621 patients were included in the present study. The study protocol was approved by the ethics committees of the participating institutions. Written informed consent was obtained from all patients.

Definition of evaluated outcomes

We evaluated two renal outcomes; worsening of the stage of diabetic nephropathy based on the level of UACR and the doubling of serum creatinine from baseline. Worsening of the stage of diabetic nephropathy was defined as transition from any baseline stage to a more advanced stage of diabetic nephropathy in two consecutive measurements of UACR. Patients who showed intermittent transition to advanced stages were not counted as worsening. This analysis was only applied to patients with normoalbuminuria and those with microalbuminuria at baseline. The other outcome was the doubling of serum creatinine from baseline. For this analysis, all patients were included regardless of the stage of diabetic nephropathy.

Statistical analysis

Data are expressed as mean ± SD or median (interquartile range). Comparisons between frequencies were tested by the χ²-test. Differences between groups were tested using the unpaired Student's t-test for normally distributed variables and Mann–Whitney U-test for variables with skewed distribution. Differences among the three groups were investigated by one-way analysis of variance (anova) for normally distributed variables and the Kruskal–Wallis test for variables with skewed distribution. Multivariate Cox proportional hazard regression model with a backward stepwise procedure was used to select those factors that significantly affected the time to the incidence of each renal outcome and to estimate the risk. The follow-up time was censored if specific renal outcome in each analysis was observed or if the patient was unavailable for follow up or experienced cardiovascular disease. Age, sex, assigned therapy arm, past history of cardiovascular disease (CVD), HbA1c, systolic BP, diastolic BP, hypertension, total-cholesterol, HDL cholesterol, triglycerides and serum uric acid were initially applied as predictive factors in the backward stepwise procedure. Continuous variables, with the exception of age, were trisected according to the number of patients and entered into the Cox model as categorical parameters. The cumulative incidence of each renal outcome was estimated using the Kaplan–Meier procedure and was compared by the log–rank test. Hypertension was defined as BP ≥ 140/90 mmHg or current use of antihypertensive drugs. All analyses were carried out by the Statistical Package for Social Sciences (version 11; SPSS, Chicago, IL, USA). A two-sided P-value of <0.05 was considered statistically significant.

Results

Clinical characteristics at baseline

The stage of diabetic nephropathy at baseline was classified as normoalbuminuria in 346 patients, microalbuminuria in 190 and overt proteinuria in 85. The mean age of all 621 participants was 71 years, and 48% of patients were men. A past history of CVD was reported by 27% of the patients. At study entry, the median HbA1c was 8.2% and mean BMI was 23.6 kg/m². Table 1 lists the baseline clinical characteristics of the study groups classified according to the assigned therapy arms. No factors at baseline were different between two groups.

Table 1.  Clinical characteristics of the study participants according to the assigned therapy arms

	Intensive therapy	Conventional therapy	P
Data are mean ± SD for normally distributed continuous variables and median(25th–75th interquartiles) for skewed continuous variables. P-values are for differences between two groups. ACR, albumin-creatinine ratio; CVD, cardiovascular disease; HDL, high-density lipoprotein; micro, microalbuminuria; normo, normoalbuminuria; NS, not significant; overt, overt proteinuria; RAS, renin–angiotensin system.
n	315	306
Age (years)	72 ± 4	71 ± 4	NS
Sex(male/female)	148/167	151/155	NS
Body mass index(kg/m^2)	23.8 ± 3.2	23.4 ± 3.3	NS
HbA1c(%)	8.2(7.9–8.7)	8.3(7.9–8.9)	NS
Systolic blood pressure(mmHg)	136 ± 17	135 ± 17	NS
Diastolic blood pressure(mmHg)	74 ± 9	76 ± 10	NS
Hypertension(%)	68	67	NS
Taking RAS inhibitors(%)	23	23	NS
Past history of CVD(%)	26	28	NS
Total cholesterol(mg/dL)	202 ± 32	203 ± 36	NS
HDL cholesterol(mg/dL)	55(46–65)	53(44–65)	NS
Triglycerides(mg/dL)	113(84–167)	110(80–159)	NS
Urinary ACR(mg/g creatinine)	20(10–96)	23(10–121)	NS
Serum creatinine(mg/dL)	0.8(0.6–0.9)	0.8(0.6–0.9)	NS
Serum uric acid(mg/dL)	4.9(4.1–5.6)	4.9(4.1–5.8)	NS
Stage of diabetic nephropathy(normo/micro/overt, %)	56 / 31 /13	56 / 30 / 14	NS

Worsening of the stage of nephropathy

Analysis of worsening of the stage of diabetic nephropathy based on the degree of albuminuria was carried out in 536 patients with normoalbuminuria and microalbuminuria at baseline. The median follow-up period was 52 months (interquartile: 28–67 months). Compared with the stage of diabetic nephropathy at baseline, 87 of 346 patients with normoalbuminuria (25%) developed microalbuminuria and 27 of 190 patients with microalbuminuria (14%) progressed to overt proteinuria. Thus, 21% of the elderly patients progressed within almost 4.5 years to a worse stage of diabetic nephropathy. Multivariate Cox proportional hazard regression analysis with backward stepwise procedure identified age, sex, and HDL cholesterol as statistically significant and independent determinants of the progression of the nephropathy stage (Table 2).

Table 2.  Factors identified by multivariate Cox proportional hazard regression analysis with backward stepwise procedure associated with worsening of the stage of diabetic nephropathy

Factor	Adjusted hazard ratio (95% CI)	P-value
Factors retained in the final model are presented. P-values were based on the Wald test, reflecting the contribution of each independent factor. HDL, high-density lipoprotein.
Age (per year)	1.05 (1.01–1.10)	0.011
Female (vs male)	1.67 (1.11–2.49)	0.013
HDL cholesterol		0.039
<49 mg/dL	1 (reference)
49 mg/dL–62 mg/dL	0.68 (0.43–1.07)	0.092
≥62 mg/dL	0.55 (0.34–0.88)	0.001

Doubling of serum creatinine

Analysis of the doubling of serum creatinine was carried out in all patients regardless of the stage of diabetic nephropathy. The median follow-up period for this analysis was 66 months (interquartile: 49–71 months). During the follow-up period, doubling of serum creatinine from that at baseline was noted in 38 patients (6.1%); including six (1.7%) with normoalbuminuria, nine (4.7%) with microalbuminuria and 23 (27.1%) with overt proteinuria. The multivariate Cox proportional hazard regression analysis identified past history of CVD, serum uric acid, and the assigned therapy group as significant and independent determinants of the doubling of serum creatinine (Table 3). In this analysis, age was retained in the final model of the backward stepwise procedure, as shown in Table 3. However, age was not identified as a significant determinant (P = 0.072). Inclusion of serum creatinine levels and the stage of diabetic nephropathy at baseline into the Cox model did not change the results.

Table 3.  Factors identified by multivariate Cox proportional hazard regression analysis with backward stepwise procedure associated with the doubling of serum creatinine

Factor	Adjusted hazard ratio (95% CI)	P-value
Factors retained in the final model are presented. P-values were based on the Wald test, representing the contribution of each independent factor. CVD, cardiovascular disease.
Age (per year)	1.07 (0.99–1.14)	0.072
Conventional therapy arm (vs intensive therapy arm)	2.03 (1.02–4.04)	0.045
Past history of CVD (vs none)	2.00 (1.04–3.84)	0.038
Serum uric acid		<0.001
<4.5 mg/dL	1 (reference)
4.5 mg/dL–6.5 mg/dL	3.42 (0.94–12.43)	0.062
≥6.5 mg/dL	8.65 (2.60–28.78)	<0.001

Effect of intervention therapy

Finally, we investigated the effects of intervention therapy on worsening of the nephropathy stage and the doubling of serum creatinine using the Kaplan–Meier method. The clinical characteristics at baseline of the two treatment groups were not significantly different (Table 1). The cumulative incidence of worsening of the nephropathy stage was similar in the treatment groups (P = 0.86, Fig. 1). As expected from the results of the multivariate Cox model, the cumulative incidence of the doubling of serum creatinine was significantly lower in the intensive therapy group than the conventional therapy group (P = 0.016, Fig. 1b).

Figure 1. Kaplan–Meier curves for cumulative incidence of (a) worsening of the stage of diabetic nephropathy and (b) the doubling of serum creatinine according to the assigned therapy arms. Differences between the groups were tested by the log–rank test. Solid line, the intensive therapy group; dotted line, the conventional therapy group.

Discussion

The present study analyzed the clinical characteristics associated with progression of diabetic kidney disease in Japanese elderly patients with type 2 diabetes. The results showed that worsening of the nephropathy stage was influenced by aging, female sex and low levels of HDL cholesterol. Furthermore, the results also showed that worsening of diabetic nephropathy represented by the doubling of serum creatinine was influenced by a past history of CVD, hyperuricemia and conventional therapy. The study also showed that patients who received intensive therapy were less likely to show doubling of serum creatinine than those assigned to the conventional therapy group, although both showed similar trends with regard to the worsening of the nephropathy stage.

In the present study, aging was identified as an independent risk factor for worsening of the nephropathy stage. Furthermore, aging contributed to the doubling of serum creatinine, although the correlation remained below the statistical significance cut-off value used in the present study. The finding is compatible with the concept that the aging process induces a variety of changes in the kidney and results in a decline in renal function.^8,9 The reported age-associated functional changes in renal function include changes in glomerular hemodynamics, decrease in renal sodium and potassium handling, and impairment of dilution and concentration ability. In addition to the aforementioned functional changes, glomerular structural changes and vascular changes have also been reported to occur with aging.¹⁰ The structural changes noted to occur with aging are very similar to those observed in the kidney of patients with type 2 diabetes. Furthermore, these age-related renal changes might accelerate in patients with type 2 diabetes. Thus, prevention of progression of diabetic nephropathy is an important therapeutic goal for all patients, which should be given attention especially for elderly patients with type 2 diabetes.

In general, the initial sign of diabetic nephropathy is an increase in urinary albumin excretion. In the present subanalysis of the J-EDIT, patients with a higher tertile of HDL cholesterol were at lower risk of progression of the nephropathy stage than those with a lower tertile of HDL cholesterol. Similar results were reported in a 7-year follow up of a type 2 diabetic cohort of Casale Monferrato, which included 1103 patients aged 68 years and older at baseline.¹¹ In the Casale Monferrato cohort, patients who developed overt nephropathy were older and had lower levels of HDL cholesterol than non-progressors.¹¹ Thus, elderly diabetic patients with lower levels of HDL cholesterol are at higher risk of increase in albuminuria, suggesting that lipid-lowering therapy and lifestyle modification that target HDL cholesterol are important to prevent the increase in albuminuria in elderly patients with type 2 diabetes.

In the present study, hyperuricemia was identified as an independent risk factor for the doubling of serum creatinine. Uric acid is reported to be only a marker of renal dysfunction but also an independent cardiovascular risk factor.^12,13 In patients with type 1 diabetes, low glomerular filtration rate was independently and significantly associated with hyperuricemia, proteinuria and old age.¹⁴ Furthermore, the increase in serum uric acid in patients with type 1 diabetes was linearly associated with the risk of early loss of renal function in a 6-year follow-up study.¹⁵ Also, recent studies found that lowering the levels of uric acid slowed the progression of renal disease. For example, Siu et al. reported that the treatment of asymptomatic hyperuricemia in patients with mild renal dysfunction (CKD stage 3) resulted in delayed disease progression.¹⁶ Taken together, the results of the aforementioned studies and those described here emphasize the need to include treatment of hyperuricemia as a therapeutic objective to prevent a progressive decline in renal function in elderly patients with type 2 diabetes.

Early decline in renal function is reported to be relatively common in diabetic patients without overt proteinuria.^17,18 In the present study, six (1.7%) patients with normoalbuminuria and nine (4.7%) with microalbuminuria showed doubling of serum creatinine. Also, patients with a past history of CVD were at risk of progressive decline in renal function. We have recently reported that abnormalities in extrarenal small vessels were a predictor for a decline in renal function in type 2 diabetic patients.¹⁹ These results highlight the importance of overall management of systemic vascular abnormalities in improving the prognosis of elderly diabetic patients by halting a progressive decline in renal function.

Interestingly, the intensive therapy group showed a lower incidence of the doubling of serum creatinine than the conventional therapy group. It is noteworthy that the baseline clinical features were similar in the two therapy groups, although no data during the follow up were available. Thus, the exact factors that contributed to the prevention of poor kidney function in the intensive therapy group remains unclear. In this regard, analysis of all participants in the J-EDIT showed no significant differences in blood pressure levels and lipid control during the follow-up period between the two groups.⁶ As well, the mean levels of HbA1c in the two groups were not different after the third year of follow-up, although the mean levels of HbA1c in the intensive therapy group decreased significantly in the first 2 years after the start of the study. Comparison of the baseline data of patients in this subanalysis and all participants in the J-EDIT showed no differences in various clinical and laboratory parameters. These results suggest that aggressive management, including those directed against multiple risk factors, might be important in the prevention of progression of renal dysfunction in elderly diabetic patients.

The present study has certain limitations. The UACR is highly variable and relatively not reproducible like serum creatinine. In the present study, we did not enrol patients whose stage of nephropathy was not identical in two assessments at baseline. These patients might rapidly progress to more advanced stages. Thus, this aspect of patient recruitment might have underestimated the predictive power of several factors on the development and progression of diabetic nephropathy. Also, we could not identify what factors were associated with lower incidence of the doubling of serum creatinine in the intensive therapy group, because clinical information including laboratory tests and medications during the follow up was not available in this analysis. However, the results stress the importance of intensive care and/or lifestyle modification for multiple risk factors in elderly patients with type 2 diabetes in preventing deterioration of renal function. Further analysis is required to clarify the effects of factors modified during the follow up and the reasonable targeted levels of these factors to prevent renal dysfunction in elderly diabetic patients.

In conclusion, our subanalysis using data of the J-EDIT showed that aging is an important factor in the progression of diabetic nephropathy in elderly patients with type 2 diabetes. The results also suggested the importance of the proper management of HDL cholesterol and uric acid in preventing the progression of diabetic kidney disease in this population. In addition, the results stress the importance of multiple risk factor intervention in preventing deterioration of renal dysfunction. To improve the prognosis of elderly patients with type 2 diabetes, it is necessary to understand their clinical features. Further accumulation of data is required to establish a proper therapeutic strategy tailored for elderly diabetic patients.

Acknowledgments

We thank all patients, physicians and staff who took part in the J-EDIT study.

The registration number for this clinical trial was UMIN000000890. This study was financially supported by Research Grants for Longevity Sciences from the Ministry of Health and Labour, and Welfare (H12-Choju-016, H15-Chojyu-016, H17-Choju-Ordinal-013) and the Japan Foundation for Aging and Health.

Conflict of interest

There is no conflict of interest. The J-EDIT Study Group has not cleared any potential conflicts.