Diabetes Care 30:2034-2039, 2007 DOI: 10.2337/dc07-0140 © 2007 by the American Diabetes Association
Normoalbuminuric Renal-Insufficient Diabetic PatientsA lower-risk group
1 Department of Nutrition-Diabétologie, Hôpital Haut-Lévêque, Pessac, France Address correspondence and reprint requests to Vincent Rigalleau, Nutrition-Diabétologie, Hôpital Haut-Lévêque, Avenue de Magellan, 33600 Pessac, France. E-mail: vincent.rigalleau{at}wanadoo.fr
OBJECTIVE—About 20% of diabetic patients with chronic kidney disease (CKD) detected from the new American Diabetes Association recommendations (albumin excretion rate >30 mg/24 h or estimated glomerular filtration rate [GFR] <60 ml/min per 1.73 m2) may be normoalbuminuric. Do the characteristics and outcome differ for subjects with and without albuminuria? RESEARCH DESIGN AND METHODS—A total of 89 patients with diabetes and a modification of diet in renal disease (MDRD) estimated GFR (e-GFR) <60 ml/min per 1.73 m2 underwent a 51Cr-EDTA B-isotopic GFR determination and were followed up for 38 ± 11 months. RESULTS—The mean MDRD e-GFR (41.3 ± 13.1 ml/min per 1.73 m2) did not significantly differ from the i-GFR (45.6 ± 29.7). Of the subjects, 15 (17%) were normoalbuminuric. Their i-GFR did not differ from the albuminuric rate and from their MDRD e-GFR, although their serum creatinine was lower (122 ± 27 vs. 160 ± 71 µmol/l, P < 0.05): 71% would not have been detected by measuring serum creatinine (sCr) alone. They were less affected by diabetic retinopathy, and their HDL cholesterol and hemoglobin were higher (P < 0.05 vs. albuminuric). None of the CKD normoalbuminuric subjects started dialysis (microalbuminuric: 2/36, macroalbuminuric: 10/38) or died (microalbuminuric: 3/36, macroalbuminuric: 7/38) during the follow-up period (log-rank test: P < 0.005 for death or dialysis), and their albumin excretion rate and sCr values were stable after 38 months, whereas the AER increased in the microalbuminuric patients (P < 0.05), and the sCr increased in the macroalbuminuric patients (P < 0.01). CONCLUSIONS—Although their sCr is usually normal, most of the normoalbuminuric diabetic subjects with CKD according to an MDRD e-GFR below 60 ml/min per 1.73 m2 do really have a GFR below 60 ml/min per 1.73 m2. However, as expected, because of normoalbuminuria and other favorable characteristics, their risk for CKD progression or death is lower.
Abbreviations: AER, albumin excretion rate CKD, chronic kidney disease e-GFR, estimated glomerular filtration rate GFR, glomerular filtration rate MDRD, modification of diet in renal disease
Twenty-five to forty percent of patients with diabetes have kidney damage (1), and diabetes is the first cause of end-stage renal disease in most countries (2). The early detection of chronic kidney disease (CKD) in diabetic patients is therefore of critical importance. The conventional approach for screening is the determination of the albumin excretion rate (AER). However, a substantial proportion of normoalbuminuric diabetic patients may present with a reduced glomerular filtration rate (GFR): their rates have been reported to be 20% based on GFR <60 ml/min per 1.73 m2 in type 2 diabetes (3) and in type 1 diabetes based on GFR <90 ml/min per 1.73 m2 with more advanced glomerular lesions (4). In accordance with the National Kidney Foundation guidelines (5), this has led the American Diabetes Association to recommend the screening of CKD in diabetic patients based both on the AER (threshold: 30 mg/24 h) and the Cockcroft and Gault equation or modification of diet in renal disease (MDRD) equation estimated GFR (threshold: 60 ml/min per 1.73 m2) (6). The MDRD equation is superior to the Cockcroft and Gault equation: it is more accurate (7,8), more robust when glucose control is poor (9), and not biased by body weight (10), which is of considerable importance because of the frequent association of type 2 diabetes, and CKD (11), with obesity. Most of the recent reports on CKD in diabetic patients rely on MDRD e-GFR (12–16). This equation may however not be ideal for screening CKD, as it is known to underestimate normal and high GFR (17,18): some of the MDRD-based CKD diagnosis may stem from this underestimation (19), particularly in normoalbuminuric patients. Furthermore, the progressive nature of normoalbuminuric CKD has been argued from studies based on measured GFR, reported by one group (3,20), and its association with complications relies on reports in Asian diabetic subjects, who are known to have a high incidence of CKD (14). Whether this applies to MDRD-diagnosed normoalbuminuric CKD in Caucasians remains to be demonstrated. To evaluate the significance of normoalbuminuric CKD in diabetes, we compared the isotopic GFR and serum creatinine of normoalbuminuric versus albuminuric diabetic patients with CKD according to an MDRD e-GFR <60 ml/min per 1.73 m2. The baseline characteristics of the patients, and their outcome during a 38-month follow-up, were also compared.
A total of 89 patients (49 men, mean age 64 ± 11 years) were recruited from the Nutrition-Diabetology and Nephrology departments of the Centre Hospitalier Universitaire de Bordeaux. The inclusion criteria were as follows: 1) diabetes (a total of 22 patients had type 1 diabetes and 67 had type 2 diabetes) and 2) CKD according to an MDRD e-GFR <60 ml/min per 1.73 m2, not requiring renal replacement therapy at inclusion. The patients gave written informed consent to participate in the study, which was approved by the ethical committee of our institution. This study was supported by a clinical research program in the Bordeaux University Hospital.
Analytical methods
Follow-up, care, and outcome
The care program objectives included glycemic control according to the French 1999 recommendations (A1C <8.0%; if possible, 6.5% without severe hypoglycemia in type 2 and <7.0% in type 1 diabetes), but also control of associated factors such as hypertension (objective: <130/80 mmHg) and dyslipidemia (objective: LDL cholesterol <1.3 g/l). We prescribed 0.8 g protein · kg–1 · day–1 according to the National Kidney Foundation recommendations (22), except for patients with clinical signs of undernutrition or who were The primary outcome was requirement for dialysis, or death. The secondary outcomes were the AER and serum creatinine for living nondialyzed patients at the end of the follow-up.
Statistical analysis
Baseline characteristics A total of 89 patients were included. Their mean diabetes duration was 18 ± 10 years, mean MDRD e-GFR was 41.3 ± 13.1 ml/min per 1.73 m2 (11–59.9), it did not significantly differ from their isotopic GFR (45.6 ± 29.7, P = 0.12), although 15 subjects had an isotopic GFR >60 ml/min per 1.73 m2. A total of 15 subjects (17%) were normoalbuminuric. The proportion of subjects whose isotopic GFR was >60 ml/min per 1.73 m2 was lower in the normoalbuminuric group (13.3%) than the normoalbuminuric group (25.7%, NS by 2). The normoalbuminuric subjects were less frequently affected by diabetic retinopathy, but they had higher HDL cholesterol and hemoglobin levels. For these characteristics, the difference was mainly due to the macroalbuminuric group, with intermediary values in the microalbuminuric group. There were also tendencies (P < 0.07) for more women, less cigarette smoking, fewer previous cardiac events, a lower duration of diabetes, and a higher total cholesterol level in the normoalbuminuric group. The type of diabetes, A1C, and blood pressure did not differ (Table 1).
Most of the normoalbuminuric patients (71%) had serum creatinine levels in the normal range. Despite their lower serum creatinine levels, their MDRD e-GFR did not significantly differ from the albuminuric group, and the MDRD underestimation was not significant in either group (P = 0.051 between isotopic and MDRD e-GFR in the microalbuminuric group). The correlation between MDRD e-GFR and the reference isotopic measurement was even better for the normoalbuminuric group, and the accuracy of the MDRD (percent estimations within ±10, 30, and 50% of the isotopic GFR) tended to be better for them. The coexistence of normal AER and CKD could therefore not be attributed to the MDRD underestimation of GFR in our patients.
Outcome
Secondary outcome. The second AER and serum creatinine determination was performed in all the alive nondialized patients. The changes in AER, serum creatinine, and MDRD e-GFR did not significantly differ according to baseline normo-, micro-, and macroalbuminuria. However, normoalbuminuria persisted during the follow-up, and serum creatinine did not change in the normoalbuminuric group. By contrast, AER increased in the microalbuminuric patients (P < 0.05 vs. baseline), and serum creatinine increased in the macroalbuminuric patients (P < 0.01 vs. baseline). The MDRD e-GFR tended to decrease in the albuminuric patients (P = 0.05), and the comparison between baseline and final values only involved the subjects who did not have to start hemodialysis (12 albuminuric patients, none normoalbuminuric) (Table 2).
Our first objective was to find out whether normoalbuminuric diabetic patients might represent an artificial group, stemming from overdiagnosed CKD due to the underestimation of high GFR by the MDRD equation. The isotopic determination of GFR in our patients argues against this hypothesis: although the MDRD underestimated their GFR, the difference was slight and not significant. The proportion of subjects for which isotopic GFR was >60 ml/min per 1.73 m2 was indeed lower in the normoalbuminuric group (13.3%) than in the normoalbuminuric group (25.7%, NS by 2), and the correlation between estimated and measured GFR was better in the normoalbuminuric group. Most of these patients, therefore, really had a GFR <60 ml/min per 1.73 m2, which would have been missed for 71% of them if their renal function had been assessed solely by a serum creatinine level >120 µmol/l. Despite its underestimation of GFR, mainly for GFR >60 ml/min per 1.73 m2 (18,23), and the increase in the number of detected CKD (12,24), the use of the MDRD equation as recommended by the American Diabetes Association is of interest, especially when AER is normal. Apart from their relatively low creatinine levels, our normoalbuminuric CKD patients displayed a number of specific characteristics: a higher proportion of women, a lower duration of diabetes, a low prevalence of retinopathy (3,4), fewer smokers (25), higher hemoglobin levels (26,27), and higher HDL cholesterol (28,29) are consistent with previous reports on the characteristics associated with the presence of an abnormal AER. As our normoalbuminuric patients were mainly women and slightly older, their low MDRD e-GFR was not unexpected: increasing age and female sex both reduce the MDRD estimation. Only a minority of these patients had diabetic retinopathy, suggesting that their renal impairment may not be due to diabetic nephropathy. However, around one-third of type 2 diabetic patients with renal biopsy–demonstrated diabetic glomerulopathy do not have retinopathy (30). Because blood pressure levels did not differ between the two groups, it seems unlikely that the low GFR of normoalbuminuric patients was due to nephroangiosclerosis; these patients were also less affected by previous cardiovascular events. However, they required as much antihypertensive therapy as did the albuminuric patients, which suggests that their renal impairment was not without consequence. Although 17 is similar to the 20 normoalbuminuric patients studied by MacIsaac et al. (3), the small sample size is a limitation of our study, which probably explains why the differences between AER and e-GFR changes did not reach significance. Formulae estimations are known to underestimate the decline in GFR in diabetic patients (31). None of the normoalbuminuric patients died or started dialysis, and their stable AER and serum creatinine, whereas AER increased in microalbuminuric, and creatinine increased in the macroalbuminuric, both significantly, are however strong indications for a better outcome in the normoalbuminuric group. It must be noticed that most of the clinical events occurred in CKD patients with macroalbuminuria. The different characteristics we found may have contributed to this better outcome: male sex (32), a longer duration of diabetes (33), the presence of retinopathy (34), a lower hemoglobin level (32,35), and smoking (36) have all been reported to be associated with the progression of nephropathy in diabetes. However, the most probable is the persisting absence of albuminuria by itself: numerous reports have emphasized its importance for CKD progression (30,33–35,37,38) and coronary heart disease (39) in diabetes. Some other limitations should be noted. As reflected by the scatter of the isotopic GFR in the albuminuric group (Table 1), their proportion of stages 4–5 CKD was higher (26/74 vs. 3/15 normoalbuminuric), which can explain that all the subjects who had to start dialysis were albuminuric. We feel nonetheless that the few severe and terminal CKD cases in the normoalbuminuric group is a reflection that it is, on the whole, a rather stable condition. Silveiro et al. (40) studied the evolution over 5 years of 51Cr-EDTA–determined GFR in 32 normoalbuminuric type 2 diabetic subjects: their GFR decline (–0.18 ml/min per month) did not differ from normal subjects (–0.14), except for 13 subjects who were hyperfiltering at baseline (–0.61). Nine of our normoalbuminuric patients were treated by ACE inhibitors or angiotensin receptor blockers. Although their proportion did not differ from the albuminuric group, these treatments may have contributed to their good outcome, but we cannot be sure of this because we do not know whether they had abnormal AER before the initiation of these medications. Finally, because no renal biopsy was performed in our patients, their renal impairment may not have been due to diabetic nephropathy, which would account for their better outcome (30). However, we feel that whether our patients actually had diabetic nephropathy is not the issue: our purpose was to determine the outcomes in patients with diabetes and a reduced estimated GFR, classified as presenting CKD according to the new recommendations. Of such patients, 17% were normoalbuminuric in our study. Our isotopic determination of GFR confirmed that they really had GFR below 60 ml/min per 1.73 m2, although most of them would have not been detected on the sole basis of their serum creatinine level. But they did not progress, and the necessity to measure estimated GFR by MDRD, and label these patients as having CKD, may result in adverse emotional and financial consequences. Further studies on the outcome of normoalbuminuric CKD in diabetes are required to demonstrate that the awareness of this condition is a benefit.
This study was supported by a grant from the Programme Hospitalier de Recherche Clinique 2001. We would like to thank Dr. Simon Jarman for revision of the English manuscript.
Published ahead of print at http://care.diabetesjournals.org on 7 May 2007. DOI: 10.2337/dc07-0140. A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C Section 1734 solely to indicate this fact. Received for publication January 23, 2007. Accepted for publication April 29, 2007.
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||