Prevention of Diabetic Nephropathy by Tight Target Control in an Asian Population With Type 2 Diabetes Mellitus: A 4-Year Prospective Analysis

Background  No study to date has evaluated whether multifactorial intervention can prevent diabetic nephropathy in patients with type 2 diabetes mellitus and normoalbuminuria. We evaluated the effect of tightly controlling multiple factors recommended by the American Diabetes Association (ADA) on the development and prevention of diabetic nephropathy in Chinese patients with type 2 diabetes mellitus and normoalbuminuria during a 4½-year period.

Methods  A longitudinal cohort study enrolled 1290 patients with type 2 diabetes and normoalbuminuria who received intensified treatment to meet the following ADA recommended goals: hemoglobin A_1c (HbA_1c), less than 7%; systolic blood pressure, less than 130 mm Hg; diastolic blood pressure, less than 80 mm Hg; low-density lipoprotein cholesterol, less than 100 mg/dL; triglycerides, less than 150 mg/dL; and high-density lipoprotein cholesterol, greater than 40 mg/dL for men and greater than 50 mg/dL for women.

Results  During the study period, 211 patients (16.4%) developed new-onset microalbuminuria. A significant association was found between the achievement of ADA goals, including HbA_1c level less than 7% (hazard ratio [HR], 0.729; 95% confidence interval [CI], 0.553-0.906; P = .03), systolic blood pressure less than 130 mm Hg (0.645; 0.491-0.848; P = .002), and high-density lipoprotein cholesterol level greater than 50 mg/dL for women and greater than 40 mg/dL for men (0.715; 0.537-0.951; P = .02) and the development of new-onset microalbuminuria.

Conclusions  Diabetic nephropathy can be delayed by tight simultaneous achievement of multiple ADA-recommended targets. This multifactorial intervention should be started in patients with diabetes and normoalbuminuria.

Diabetic nephropathy (DN) is the leading cause of end-stage renal disease (ESRD) worldwide.¹ The prevalence of DN is greater in patients with type 2 diabetes mellitus of many ethnicities, including black, Hispanic, and Asian, with the greatest prevalence found in Asians.^2,3 When we compared international data using the US Renal Data System, we found Taiwan to have had the highest incidence of ESRD and the second highest prevalence in the world from 2002 to 2005.^4,5 According to Yang and Hwang,⁶ the increased prevalence of DN in Taiwan is the main cause of the increases in prevalence and incidence of ESRD there.

Microalbuminuria is reported to be a powerful independent risk factor for DN in patients with type 2 diabetes.^7,8 Albuminuria has also been associated with increased risk of all-cause and cardiovascular mortality in this patient group.^9-11 In a recent study by some of us,¹² an association was also found between patients with diabetes and albuminuria and coronary artery disease. Therefore, the control of microalbuminuria may halt progress to overt nephropathy and reduce occurrence of cardiovascular events in these patients.¹³ Albuminuria in patients with type 2 diabetes has been found to be predicted by increased systolic blood pressure (SBP), cholesterol level, and urinary albumin excretion rate and worse glycemic control by previous studies,^14-18 all of which analyzed the relationship between baseline data and incidental albuminuria. To our knowledge, no study has analyzed longitudinal data to identify risk factors for new-onset microalbuminuria. The Steno-2 Study has already demonstrated that intensified treatment results in a significant decrease of DN in patients with diabetes and microalbuminuria.^19,20 However, no study has evaluated the effect of multifactorial intervention on the development of microalbuminuria in patients with diabetes and normoalbuminuria.

This study evaluated the effect of tightly controlling multiple factors recommended as targets by the American Diabetes Association (ADA) on the development and prevention of DN in a cohort of patients with normoalbuminuria and type 2 diabetes. We did so by measuring hemoglobin A_1c (HbA_1c) levels, blood pressure, and lipid profiles every 3 to 6 months during a 4-year period.

This study included all patients with type 2 diabetes who visited the diabetic clinic in the Division of Endocrinology and Metabolism at Kaohsiung Medical University Hospital and Changhua Christian Hospital from January 1, 2004, through April 31, 2005. The follow-up period lasted until August 31, 2008. To be included in this study, patients (1) had to have been diagnosed as having type 2 diabetes and been followed up regularly every 3 to 6 months in the outpatient departments of the hospitals and (2) had to have normoalbuminuria (diagnosed as 2 urinary albumin-creatinine ratios [ACRs] <30 mg/g) and normal plasma creatinine levels (≤1.5 mg/dL for men and ≤1.4 mg/dL for women; to convert creatinine to micromoles per liter, multiply by 88.4). We excluded patients who had acute myocardial infarction, heart failure, or stroke within the 1-year period leading up to their inclusion into this study. The protocol for this study was approved by the human research ethics committees at Changhua Christian Hospital and Kaosiung Medical University Hospital; informed consent was obtained from each patient.

For each patient, we conducted an interview and a comprehensive assessment of disease status, complications, and risk factors, including blood and urine analyses, calculated body mass index (BMI; calculated as weight in kilograms divided by height in meters squared), and mean blood pressure (including SBP and diastolic blood pressure [DBP]) on the basis of 2 measurements taken from patients in seated positions after 10 minutes of rest. All blood samples were obtained at 8 AM after an overnight fast and before morning medication. Blood samples were analyzed for plasma glucose, total cholesterol, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglycerides, and creatinine using a biochemistry automatic analyzer (Beckman Coulter Inc, Fullerton, California). The HbA_1c was measured in whole blood using ion exchange high-performance liquid chromatography (Variant II Turbo; Bio-Rad Laboratories, Hercules, California).

All patients were followed up regularly every 3 to 6 months. At each follow-up, patients received the same tests they had received at baseline. It was requested that all patients take part in a strict intervention program aimed at meeting the treatment goals recommended by the ADA. Individualized nutrition plans were prescribed on the basis of recommendations made by the ADA and were adjusted on the basis of patient preferences, ideal body weight (ie, healthy BMI: in Taiwan, less than or equal to 24), and endemic, religious, and socioeconomic factors. The ADA suggestions for physical activity recommend at least 150 minutes per week of aerobic exercise and no more than 2 consecutive days without exercise. A physician, nurse, and dietitian worked as a team to modify patient behavior through counseling and educational programs to meet the following ADA-recommended goals: HbA_1c, less than 7%; SBP, less than 130 mm Hg; DBP, less than 80 mm Hg; LDL-C, less than 100 mg/dL (to convert LDL-C to millimoles per liter, multiply by 0.0259); triglycerides, less than 150 mg/dL (to convert triglycerides to millimoles per liter, multiply by 1.8); and HDL-C, greater than 40 mg/dL for men and greater than 50 mg/dL for women (to convert HDL-C to millimoles per liter, multiply by 0.0259). Statins were prescribed to patients with LDL-C levels above 100 mg/dL, and fibrates were prescribed to patients with triglyceride levels above 150 mg/dL or HDL-C levels under 40 mg/dL in men and under 50 mg/dL in women.

We collected overnight first-void urine samples and fasting venous blood samples from each patient to measure urine albumin and creatinine. Using these 2 measurements, we calculated an ACR for each patient. Normoalbuminuria was defined as an ACR of less than 30 mg/g. Patients were considered to have microalbuminuria if their ACR ranged from 30 to 299 mg/g in at least 2 of 3 tests they received during the 6-month study period. Patients were considered to have overt proteinuria if they had an ACR of 300 mg/g or higher. We estimated the glomerular filtration rate (GFR) using the equation recommended by the National Kidney Foundation in the Modified Diet in Renal Disease.²¹ The 24-hour creatinine clearance was corrected for a standard body surface area of 1.73 m².

New-onset microalbuminuria was defined as having an ACR ranging from 30 to 299 mg/g in at least 2 of 3 tests during a period of 6 months. If a patient was found to have an ACR ranging from 30 to 299 mg/g, then he or she would undergo 2 additional ACR measurements every 3 months. The start date for new-onset microalbuminuria was the date the first elevated ACR measurement was found.

The baseline clinical and biochemical features of the population are presented as mean (SD) or percentages. The number of follow-ups for each participant ranged from 9 to 23 (16.1). A mean value for each variable (BMI, waist circumference, SBP, DBP, HbA_1c, fasting glucose, total cholesterol, triglycerides, LDL-C, HDL-C, and estimated GFR) was calculated on the basis of all follow-up data each patient had. These values (follow-up laboratory values) were then grouped by new onset of microalbuminuria status and reported as mean (SD). Because the distributions of triglycerides and urinary ACR were highly skewed, these 2 variables were natural log–transformed in our analyses. The t test was used to compare continuous variables between groups, and χ² tests were used to compare categorical variables. The paired t test was used to compare the mean difference between baseline and end of follow-up data. Cox regression analysis was used to examine the variables that might predict new-onset microalbuminuria. Results are described as hazard ratios (HRs) and their 95% confidence intervals (CIs). Multiple comparisons for new-onset microalbuminuria analyses were performed by means of Bonferroni correction (Figure). The P value was multiplied by the number of comparisons performed. All statistical operations were performed by means of the SPSS statistical package, version 14.0 (SPSS Inc, Chicago, Illinois).

Initially, 1303 patients were enrolled in the study. Thirteen patients were lost to follow-up during the initial 6 months and excluded from our study. Therefore, this study included a total of 1290 patients with diabetes and normoalbuminuria (578 men and 712 women). The mean age of the cohort was 62.9 years, the mean duration of diabetes was 10.1 years, and the mean follow-up period was 3.8 years. On average, all patients had a mean (SD) of 16.1 (3.3) visits during the study.

Clinical and biochemical characteristics of the patients at baseline and at the end of the trial are given in Table 1. The blood pressure and fasting glucose, HbA_1c, total cholesterol, triglyceride, and LDL-C levels were significantly lower and the HDL-C levels significantly higher at the end of study compared with those at baseline. More patients were receiving insulin therapy, antihypertensive treatment, and lipid-lowering treatment by the end of the study.

During the study period, 211 patients (16.4%) developed new-onset microalbuminuria. Table 2 summarizes the baseline demographic, clinical, and biochemical characteristics of patients who did and did not later develop new-onset microalbuminuria during the follow-up period. Patients with diabetes who developed new-onset microalbuminuria were more likely to be older, have a longer duration of diabetes, have higher SBP and ACR, and have lower HDL-C level and estimated GFR at baseline than those who did not. The socioeconomic status was not significantly associated with new onset of microalbuminuria. Multiple logistic regression analysis revealed only baseline estimated GFR to be significantly associated with later development of new-onset microalbuminuria.

Patients who developed microalbuminuria had significantly higher mean SBP and fasting plasma glucose levels than those who did not develop microalbuminuria (Table 3). The mean HDL-C levels were significantly lower in patients who developed microalbuminuria than in those who did not.

The relationship between attainment of treatment goals and new-onset microalbuminuria was analyzed (Table 4). In the Cox regression model, a follow-up SBP less than 130 mm Hg (HR, 0.65; 95% CI, 0.49-0.85; P = .002), HbA_1c level less than 7% (0.73; 0.55-0.91; P = .03), and HDL-C level greater than 50 mg/dL in women and greater than 40 mg/dL in men (0.72; 0.54-0.95; P = .02) were significantly associated with decreased incidence of microalbuminuria after adjusting for sex, age, medications (antihypertensive and lipid-lowering drugs), and BMI (Table 4). The DBP, LDL-C, and triglyceride goals were not found to be associated with onset of microalbuminuria. The risk association of new-onset microalbuminuria was attenuated after adjustment for sex, age, medications (antihypertensive and lipid-lowering drugs), and baseline variables (BMI, SBP, DBP, HbA_1c, total cholesterol, triglycerides, LDL-C, HDL-C, creatinine, duration of diabetes, ACR, and estimated GFR). A follow-up SBP of less than 130 mm Hg (HR, 0.74; 95% CI, 0.56-0.98; P = .03) was significantly associated with decreased incidence of microalbuminuria.

Patients were stratified according to the number of target goals they met during the study period: HbA_1c less than 7%, SBP less than 130 mm Hg, and HDL-C greater than 50 mg/dL for women and greater than 40 mg/dL for men. During the study period, 104 patients (8.1%) attained 3 targets, 921 (71.4%) attained 1 or 2 targets, and 265 patients (20.5%) did not attain any goal. The Figure shows that the fewer goals attained, the greater the probability of developing new-onset microalbuminuria (P = .002). Patients with diabetes who attained 2 or 3 targets had a significantly lower cumulative incidence of new-onset microalbuminuria than those who did not attain any goal (P < .001). Patients who attained 1 target had an insignificantly lower cumulative incidence than those did not attain any goal (P = .35). At the time of enrollment, 53 patients (4.1%) had attained 3 targets, 946 (73.3%) had attained 1 or 2 targets, and 291 (22.6%) did not attain any goal. No association was found between the number of targets achieved at baseline and new onset of microalbuminuria (P = .61).

Two hundred seventeen patients had at least 1 episode of hypoglycemia during the study period. Four patients had major hypoglycemia without clinical morbidity and mortality. Twelve had symptomatic hypotension, which was relieved by adjustment of medications. A total of 225 patients had cardiovascular diseases. During the study period, 37 patients died.

Our study found no association between baseline HbA_1c level, blood pressure, or lipid profiles and risk of microalbuminuria. We found a significant association between being able to closely control ADA goals, including glucose level (HbA_1c level <7%; HR, 0.73; 95% CI, 0.55-0.91; P = .02), blood pressure (SBP <130 mm Hg; 0.65; 0.48-0.85; P = .002), and HDL-C level (HDL-C level >50 mg/dL in women and >40 mg/dL in men; 0.72; 0.54-0.95; P = .02) and the development of new-onset microalbuminuria. Our results also found a significant association between tight simultaneous achievement of multiple ADA-recommended targets during the study period, but not at baseline, and the reduction of new-onset microalbuminuria. By using continual patient education and careful attention to maintaining ADA-recommended target goals, we could reduce the risk of DN in patients with type 2 diabetes.

This longitudinal study is the first, to our knowledge, to evaluate multifactorial interventions on the development of new-onset microalbuminuria in type 2 diabetes with normoalbuminuria. By the end of the study period, our patient population had a mean HbA_1c level of 7.3%, an SBP of 129.3 mm Hg, a DBP of 74.4 mm Hg, an LDL-C level of 98.6 mg/dL, a triglyceride level of 116.0 mg/dL, and an HDL-C level of 53.6 mg/dL (Table 1). However, even with close attention, not all our patients could achieve the ADA-recommended goals. Those who reached them had a significantly lower rate of new-onset microalbuminuria than those who did not. Our findings suggest that such intervention can be used at the very early stages of diabetic renal disease and that such early intensive multiple interventions can reduce the risk of DN.

The SBP of a patient throughout the trial, not baseline blood pressure, was found to be significantly associated with increased risk for new-onset microalbuminuria. The SBP appeared to be the strongest risk predictor of the development of microalbuminuria, and its reduction was found to be the most protective factor against it. This finding is consistent with that of a previous study²² that reported SBP was a better predictor of microalbuminuria in patients with normoalbuminuria than DBP and a study²³ that reported high SBP was significantly associated with progression of renal disease in diabetic patients with overt nephropathy. The UK Prospective Diabetes Study²⁴ (UKPDS) found that intensified blood pressure control failed to produce a significant reduction of microalbuminuria. However, this failure of intensified blood pressure control in the UKPDS to prevent microalbuminuria might have been related to the fact that a considerable portion of that cohort had blood pressure higher than the ADA-recommended target. Recently, Ruggenenti et al²² reported finding an independent association between reduced blood pressure and risk of microalbuminuria. They reported that patients with diabetes and follow-up SBPs less than 139 mm Hg were at significantly less risk for microalbuminuria than those who had SBPs of 139 mm Hg or higher. Our study found that DN could be prevented by maintaining an SBP of less than 130 mm Hg throughout the trial.

Our results also found that new-onset microalbuminuria could be reduced by strict glucose control (HbA_1c level <7%). Other studies have had similar findings. For example, the UKPDS reported that tighter glucose control reduced the development of microalbuminuria throughout a prolonged follow-up period.²⁵ The recently published Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation (ADVANCE) trial²⁶ reported that intensive glucose control that lowered the HbA_1c values to 6.5% yielded a 21% relative reduction in nephropathy and a 9% reduction in new-onset microalbuminuria, although the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial reported increased mortality in patients receiving intensive glucose therapy targeting HbA_1c levels. Recently, the ADA stated that the general goal of an HbA_1c level less than 7% appears reasonable for patients with type 2 diabetes.²⁷ On the basis of these findings, the kind of strict glycemic control (HbA_1c level <7%) we used in this study may help prevent DN in patients with diabetes.

The association between dyslipidemia and albuminuria has been discussed in previous studies, although conclusions have been inconsistent. Serum cholesterol level has been associated with the development of increased urinary albumin excretion in Pima Indians,¹⁴ and baseline serum cholesterol level has been found to be an independent risk factor for DN in patients with type 2 diabetes.¹⁶ In a study of 90 patients with type 2 diabetes, Kim et al²⁸ reported triglycerides to be a factor in the progression of DN. Similarly, the UKPDS reported fasting plasma triglyceride level to be a strong independent determinant of microalbuminuria and macroalbuminuria. In prospective studies¹⁶ of patients with type 2 diabetes, an elevated triglyceride-to–HDL-C ratio has been independently associated with the progression of microalbuminuria. The differences in the results of these studies may have resulted from differences in sample sizes, ethnic groups, or study designs (cross-sectional or longitudinal). We found no association between baseline cholesterol, triglyceride, HDL-C, or LDL-C levels and new-onset microalbuminuria. Throughout the trial, cholesterol, triglyceride, or LDL-C levels could not be associated with new-onset microalbuminuria. The reason that lowering of LDL-C levels did not have an effect on reduction of new-onset microalbuminuria may have been because there was only a small reduction in LDL-C level (103.5 mg/dL at baseline and 98.6 mg/dL at the end of study). We found an association between mean follow-up HDL-C level and new-onset microalbuminuria. Similarly, Xu et al²⁹ reported low HDL-C level to be a predictor of incidental albuminuria in a longitudinal study.

This study has several limitations. First, we did not examine the contribution of genetic factors, which are complex and not routinely measured in daily clinical practice. Second, glucose, blood pressure, and lipid profiles are all continuous variables and the target cutoff points might be arbitrary. However, the international and regional guidelines all recommend treatment targets for diabetic management. These treatment targets are common and their usage easy for physicians. Another study limitation was that we did not include a comparison group. This study was designed to evaluate the effect of tightly controlling multiple factors recommended as targets by the ADA on the development and prevention of DN in a cohort of patients with normoalbuminuria and type 2 diabetes. The previous studies revealed that it is hard for all patients in the intensive-treatment groups to achieve the treatment goals. Even with close attention, not all our patients could achieve the ADA-recommended goals. Those who reached them had a significantly lower rate of new-onset microalbuminuria than those who did not. Despite these limitations, this cohort study had a relatively large sample size, detailed phenotyping, and a relatively long observational period (mean, 4 years). It was a longitudinal study on the basis of data measured at more than 2 different times and used methods for analyzing longitudinal data and multivariable-adjusted analyses. Moreover, this study defined DN on the basis of the ADA consensus guideline that recommends at least 2 urine samples be collected during a 3- to 6-month period to appropriately classify patients as having normoalbuminuria, microalbuminuria, or macroalbuminuria.

In conclusion, this study found a strong association between tight simultaneous control of multiple ADA-recommended target factors and decreased risk of DN. This multifactorial intervention should be started in patients with diabetes and normoalbuminuria.

Correspondence: Ming-Chia Hsieh, MD, PhD, Divisions of Endocrinology and Metabolism, Department of Internal Medicine, Kaohsiung Medical University Hospital, Kaohsiung, Taiwan (mingchia0531@yahoo.com.tw).

Accepted for Publication: September 10, 2009.

Author Contributions:Study concept and design: Tu, Chang, J.-F. Chen, Tien, Hsiao, H.-C. Chen, and Hsieh. Acquisition of data: Tu, Chang, J.-F. Chen, Tien, H.-C. Chen, and Hsieh. Analysis and interpretation of data: Tu, Chang, J.-F. Chen, Hsiao, and Hsieh. Drafting of the manuscript: Tu and Hsieh. Critical revision of the manuscript for important intellectual content: Tu, Chang, Tien, Hsiao, H.-C. Chen, and Hsieh. Statistical analysis: Tu, Chang, and Hsieh. Obtained funding: Tu, J.-F. Chen, Tien, Hsiao, and Hsieh. Administrative, technical, and material support: Tu, Chang, J.-F. Chen, Tien, Hsiao, H.-C. Chen, and Hsieh. Study supervision: Tu, Chang, J.-F. Chen, Hsiao, H.-C. Chen, and Hsieh.

Financial Disclosure: None reported.