- Email: [email protected]
Retrospective analysis of hemodialyzed diabetic patients in Japan
ELSEVIER
Retrospective analysis of hemodialyzed diabetic patients in Japan Izumi Takei*, Akira Yamauchi, Shinya Nakamoto, Hiromichi Suzuki, Takao Saruta Department
Of
Internal Medicine, Keio University School
of
.&f&c&,
Shinanomachi
35,
Shinjtku-ku,
T&o,
Japn
160
Received9 May 1995;revision received3 August 1995;accepted8 August 1995
Abstract We retrospectively analyzed the courses of 37 non-insulin dependent diabetics (hemodialyzed:HD group) with endstagerenal disease(ESRD), to identify factors predisposing to renal failure. The factors analyzed were: diabetic (nonproliferative and proliferative) retinopathy, family histories of diabetes and hypertension, smoking, dyslipidemia, first examination proteinuria and non-compliance. These factors were statistically compared in 37 NIDDM without renal failure (non-HD group). There were no significant differences in age or duration of diabetes between the two groups. Significant differences (P < 0.001) were, however, recognized in diabetic proliferative retinopathy and hypertension betweenthe two groups. Hypertension was present in 35/36(97.2%) HD patients and in 21/36(58.3%)non-HD patients. A family history of hypertension was recognized in 16/37HD (43.2%) and in 7/33 (21.2%) non-HD (P < 0.05). Differenceswere recognized in HDL-cholesterol, LDL-cholesterol and TG levels (38.2 l 12.5mg/dl and 56.7 f 18.5mg/dl, 140.4 f 57.1 mg/dl and 115.6 f 33.6 mg/dl, 169.9 f 89.4 mg/dl and 115.7 f 75.1 mg/dl, in HD and non-HD, respectively, P c 0.05). First visit proteinuria was found in all HD patients, and in 6/34 (17.6%) non-HD. The difference in previous treatment refusal, for 7 or more years, was significant with 23/36 (58.9%) HD patients and only l/25 (4.0%) non-HD patients (P < 0.001) having a history of prolonged non-compliance with diabetic treatment. Diabetic retinopathy, non-proliferative and proliferative, hypertension and a family history of hypertension, elevated triglyceride and LDL-cholesterol, low HDL-cholesterol, first visit proteinuria, and prolonged non-compliance correlated with progression to ESRD. We advocate expanding diabetic education to include prevention of complications such as diabetic nephropathy. Keywords: Non-insulin dependent diabetes; End stage renal disease; Hemodialyzed; First visit proteinuria; Noncompliance
1. Introduction
The incidence of renal failure, due to diabetic nephropathy, has increased dramatically in recent l Corresponding author, Tel.: +81 3 33531211 Ext. 2383; fax: +81 3 33592745.
decades. In Japan, the majority of end-stage renal
disease(ESRD) patients are managed with hemodialysis (HD). In addition to the considerable financial burden on medical facilities, HD is associatedwith increased morbidity and mortality in diabetic patients [l-3]. Although
0 1995ElsevierScienceIreland Ltd. All rights reserved 0168~8227/95RO9.50 SSDI 0168-8227(95)01130-6
insulin
dependent diabetics are at
174
I. Takei et al. /Diabetes
Research and Clinical Practice 29 (1995) 173-I 77
higher risk of developing nephropathy and of progressing to ESRD, diabetic nephropathy is also a common finding in those with non-insulin dependent diabetes mellitus (NIDDM) [4]. Reported rates of ESRD in NIDDM patients with nephropathy range from 3- 10% [5-71. As NIDDM is the more common type of diabetes, particularly in Japan, these patients represent the majority of the HD population. It is widely recognized that appropriate medical therapy, especially tight glycemic control, can prevent or delay the development of diabetic nephropathy. Even in those with significant proteinuria (albumin excretion rate > 250 &min), which may occur several years after the diagnosis of NIDDM, loss of renal function can be decelerated with aggressive treatment [8]. Diabetic nephropathy, one of the most serious complications of diabetes, is clinically silent for lo- 15 years. The slow development and long latency period of diabetic nephropathy, while providing ample opportunity for medical intervention, are also major sourcesof patient noncompliance. The problem of noncompliance is one of particular significance in Japan. Only a few decades ago, diabetes was an uncommon diseasein Japan and of little concern to the averageperson. Since the end of the second world war the prevalence of diabetes, especially NIDDM, has increased dramatically. Most Japanesediabetologists attribute this remarkable rise to life-style, particularly dietary, changes. The increased prevalence of NIDDM in our society has not, however, led to increased public awarenessof its complications. The need for prevention of diabetic nephropathy, as well as other serious complications such as retinopathy and neuropathy, is clear. Thus, with the aim of identifying risk factors for the development of diabetic nephropathy, we retrospectively analyzed 37 NIDDM patients on HD. Poor glycemic control, hypertension, a family history of hypertension, smoking, dyslipidemia, and microalbuminuria, as well as the presenceand severity of other diabetic complications, have all been reported to be associated with the development of diabetic nephropathy [2,9,10]. Several of these factors, believed to influence the progression of diabetic nephropathy, were examined in the present study.
2. Subjects and methods The study group consisted of 37 NIDDM patients who had undergone HD at Keio University Hospital within the 5 year period from 1988 through 1992.There were 24 males, 13 femalesand the average age was 62.0 f 10.0 years. The average duration of diabetes before the induction of HD was 17.4 f 7.0 years. The average HbA,c percentagein the 6 months before the induction of HD was 7.0 i 0.9. The control group was made up of 37 NIDDM patients who had never been on HD. There were 19 males and 18 females, with an average age of 65.5 f 11.6years, all of whom had been followed at our out-patient clinic for more than 10 years. All were being treated with insulin or oral hypoglycemic agents. The non-HD and HD groups were well matched in terms of diseaseduration, the average duration of diabetes in the nonHD group being 18.7 f 6.9 years. The average
Table 1 Comparison between the HD and the non-HD groups HD Number of subjects (male/female) Age (mean f SD.) Disease duration (years) HbA,c (%) Complications Proliferative retinopathy Non-proliferative retinopathy No retinopathy Hypertension Family history of hypertension Family history of diabetes mellitus Smoking
non-HD
&3) &8) 62.0 f 10.0 65.5 zt 11.6 17.4 f 7.0 18.7 f 6.9 7.0 l 0.9 8.6 zk l.7* 30136 (83.3%) 6136 (16.7%) O/36 We/,) 35136 (97.2%)
6136 (16.7%)’ 16136 (44.4%)** 14136 (38.9%)* 21136 (58.3%)*
16137 (43.2%) 17136 (47.2%) 15/3l (48.4%)
7133 (21.20/o)*** 21133 (63.6%) lU33 (36.4%)
*P c 0.001; **p c 0.01; l **P c 0.05.
1. Takei et al. /Diabetes
Research and Clinical Practice 29 (1995) 173-177
HbAic percentage in the 6 months prior to this study was 8.6 f 1.7. We periodically measured average lipid metabolic characteristics, specifically total cholesterol (TC), HDL-cholesterol (HDL-C), LDL-cholester01 (LDL-C), and triglycerides (TG), in the HD-group for the 6 month period before the induction of HD and also made thesedeterminations in the non-HD group 6 months prior to starting the study. The results of the two are compared in Table 1. Thirty-six subjects from each group were examined for the existence of diabetic retinopathy, hypertension, family histories of hypertension and diabetes,and smoking, The prevalence of first visit proteinuria, the time course of renal function, and the frequency and duration of periods of noncompliance with treatment were also assessed. The statistical differences were assessed by Fisher’s exact test and non-parametric analysis (Mann Whitney U-test). The data are expressedas means f S.D. A value of P < 0.05 was considered to be statistically significant. 3. Results
Significant differences (P < 0.01) between the HD and non-HD groups were recognized in diabetic retinopathy, diabetic proliferative retinopathy and hypertension. Non-proliferative diabetic retinopathy was identified in 6 out of 36 HD patients (16.7%) and in 16 out of 36 non-HD patients (44.4%), who underwent ophthalmological examination. A significant difference was seen in the prevalence of proliferative diabetic retinopathy, which affected 30 HD patients (83.3%) but
175
only 6 non-HD patients (16.7%) (P < 0.001). Fourteen non-HD patients had no signs of diabetic retinopathy (38.9%). Hypertension was a common finding in both groups, affecting 35 of 36 HD (97.2%) and 21 out of 36 non-HD (58.3%) patients (P < 0.001 ) (Table 1). Detailed family histories were obtained from all of the HD study participants and from 33 non-HD subjects. It was found that 17 of the 36 HD patients (42.2%) and 21 of the 33 non-HD patients (63.3%) had a family history of diabetes, but the difference did not reach statistical significance. A marked difference was, however, recognized for hypertension with 16 of 37 (43.2%) HD but only 7 of 33 (21.2%) non-HD patients reporting a family history of high blood pressure (P < 0.05). Numbers of smokers, 15 of 31 HD and 12 of 33 non-HD patients, were similar in the two groups (Table 1). However no attempts were made to distinguish among light, moderate and heavy smokers or among smoking durations. Those who had quit smoking in the past were considered nonsmokers at the time of the study. Total cholesterol, HDL-cholesterol and serum triglyceride (TG) levels were determined as parameters of lipid metabolism (Table 2). Significant differences were observed in HDL-cholesterol, LDLcholesterol and TG levels. The HDLcholesterol levels were 38.2 f 12.5 and 56.7 f 18.5 (P < O.OOl), LDL-cholesterol levels 140.4 f 57.1 and 115.6 f 33.6 (P < 0.01) and the
I ,CRN
Table 2 Lipid metabolic parameters HD TC (mg/dl) HDL-C (mg/dl) LDLC(mg/dl) TG (mg/dl)
210.7 38.2 140.4 169.9
non-HD zt zt f *
61.6 12.5 57.1 89.4
195.5 56.7 115.6 115.7
f f f f
O!
43.8 18.5* 33.6*** 75.1**
(meanf
S.D.) l P < 0.001; l *P < 0.01; l **p < 0.05.
Fig. 1. Tie
course of renal function (lkxum
creatinine level).
176
1. Takei et al. / Diabeles Research and Chtical Pracfice 29 (1995) /73- 177
Table 3 First visit proteinuria and non-compliance
Proteinuria Non-compliance (7- 11 years without treatment)
HD
non-HD
36/36 (100%) 22/36 (61.1%)
6134(17.6%)* l/25 (4.00/O)*
‘P c 0.001.
serum TG levels 169.9 f 89.4 and 115.7 * 75.1 (P < O.Ol), in the HD and non-HD groups, respectively. The average total cholesterol level 6 months prior to dialysis was higher in the HD group (210.7 f 61.6 as compared to 195.5 f 43.8 in the non-HD group) but the difference was not statistically significant (Table 2). Review of the medical records revealed that proteinuria, of some degreehad been present in all 36 HD patients tested at the initial visit to our outpatient clinic. In contrast, only 6 of the 34 non-HD patients tested had presented with proteinuria (P < 0.001). The reciprocal of the serum creatinine level was used to assessthe time course of renal function. As shown in Fig. 1, over a follow-up period of 70 months, this index fell sharply in the HD group. Among the non-HD patients who had initially presented with proteinuria, there was a slight diminution which correlated with the duration of diabetes. The l/Cm value remained stable in the non-HD patients who did not have proteinuria. The most striking difference identified in this study was in the numbers of patients who had previously refused treatment, for periods ranging 7-l 1 years. An extended period of non-compliance with diabetic therapy and management was reported by all 36 (100%) HD but only 1 of the 25 (4.0%) non-HD patients (P < 0.001 ) (Table 3). 4. Discussion
We conducted a retrospective analysis of HD and non-HD NIDDM patients, for the purpose of identifying risk factors for the development of diabetic nephropathy. As expected, microvascular
complications were present in the majority of those who had been diabetic for more than 10 years. However, the prevalences of both simple and proliferative retinopathy were significantly higher in HD than in non-HD patients. The HD group was also found to have markedly higher rates of hypertension and a family history of hypertension. All but one of the HD patients was hypertensive, and most were treated with calcium channel blockers. As demonstrated by a number of studies in recent years, disorders of lipid metabolism, predisposing to cardiac death, are common fmdings in diabetics with ESRD due to diabetic nephropathy. Our results are in accordance with previous studies concerning association between dyslipidemia and diabetic nephropathy. Levels of HDL-cholesterol were significantly higher, those of TG significantly lower, in the HD than in the non-HD group. The establishment of lipid parameters as predictors of progression to ESRD awaits further clarification of the possible contributions of lipids to the development of diabetic nephropathy. All of the HD patients had initially presented with proteinuria, while less than one fifth of the non-HD patients had been proteinuric at the first visit to our out-patient clinic. Renal function, as assessed by taking the reciprocal of serum creatinine clearance, declined markedly over the following 70 months in the HD group. In contrast, renal function remained relatively stable in the non-HD group. Among the factors examined in this study, neither smoking nor a family-history of diabetes was found to be significantly related to the development of diabetic nephropathy and progression to ESRD. The role of cigarette smoking in the development of albuminuria, reported by Chase et al., may pertain only to insulin-dependent diabetics [13]. The influence of family history, i.e. a genetic predisposition to the development of diabetic nephropathy, has also been investigated primarily in insulin-dependent diabetics [ 14- 161. The role of these factors, if any, in nephropathy and ESRD in NIDDM await further investigation. The striking correlation between extended periods of non-compliance with treatment and
I. Takei et al. /Diabetes
Research and Clinical Praclice 29 (1995) 173-I 77
progression to ESRD has significant educational, as well as prognostic, significance. In recent years, several investigations have emphasized that NIDDM patients are at risk for developing many of the same renal complications as those with IDDM [4,10,17,18]. The risk of developing diabetic nephropathy, with the possibility of progression to ESRD has not been fully appreciated by either patients or clinicians. Diabetes was a relatively uncommon diseasein Japan until 15-20 years ago. As NIDDM is often asymptomatic for several years, many patients diagnosed in the 1970sand 1980sdid not consider it to be a serious health problem. More than half of the HD patients, in the present study, reported having refused treatment for periods of 7-l 1 years. In contrast, only one of the non-HD patients had done so. NIDDM is not simply a very mild form of IDDM. Newly diagnosed NIDDM patients must be fully appraised as to the nature and long-term significance of their disease. Diabetic education, aimed at preventing or slowing the development of complications, is urgently needed. A full understanding of the risks inherent to non-compliance, especially of poor glycemic control and hypertension, ‘will motivate patients to obtain adequate medical care. The prevalence of NIDDM is still rising in Japan. It is not uncommon for proteinuria to be present at, or shortly after, the time of diagnosis. Treatment must not be delayed in such cases. We therefore recommend that diabetic education, aimed at newly diagnosed NIDDM patients, be instituted at the first clinic visit. References [l]
Ghavamian, M., Gutch, C. and Koop, F. (1982) The sad truth about hemodialysis in diabetic nephropathy. J. Am. Med. Assoc. 222, 1368-1369. [2] Matson, M. and Kjellstand, CM. (1988) Long-term follow-up of 369 diabetic patients undergoing dialysis. Arch. Intern. Med. 148, 600-604. (31 Bell, D.S.H. (1991) Diabetic nephropathy: changing concepts of pathogenesis and treatment. Am. J. Med. Sci. 301, 195-200.
117
predicts clini141 Mogensen, C.E. (1984) Microalbuminuria cal proteinuria and early mortality in maturity-onset diabetes. N. Eng. J. Med. 310, 356-360. 151Schimitz, A. and Vaeth, M. (1988) A major risk factor in non-insulin dependent diabetes. A IO-year follow up study of 503 patients. Diabetic Med. 5, 126-134. 161Reddi, A.S. and Camerini-Davalos, R.A. (1990) Diabetic nephropathy: an update. Arch. Intern. Med. 150, 31-43. W.G. (1991) Diabetic I71 Sirmon, M.D. and Kirkpatrick, nephropathy: new direction in management. Renal Failure 13, 51-59. 181Mogensen, C.E. (1976) Progression of diabetic nephropathy in long-term diabetics and effect of initial antihypertensive treatment. Stand. J. Clin. Lab. Invest. 36, 383-388. 191 Gomi, Y., Suzuki, S., Iino, S. et al. (1992) The possibility of hereditary factors in the susceptibility to diabetic nephropathy in NIDDM. Folia. Endocrinol. 68, 592-599. 1101Ritz, E., Raine, A. and Cordonnie, D. (1994) Hemodialysis in Type I and Type II diabetic patients with end stage renal failure. In: C.E. Mogensen (Ed.), The Kidney and Hypertension in Diabetes Mellitus, Kluwer Academic Publishers, Dordrecht, pp. 459-467. 1111Aurell, M. and Bjorck, S. (1992) Determinants of progressive renal disease in diabetes mellitus. Kidney. lnt. Suppl. 36, S38-S42. 1121Tschope, W., Koch, M., Thomas, B. and Ritz, E. (1993) Serum lipids predict cardiac death in diabetic patients on maintenance hemodialysis. Nephron 64. 354-358. 1131 Chase, S.P., Garg, S.K. and Hamman, R.E. et al. (1991) Cigarette smoking increases the risk of albuminuria among subjects with type I diabetes. J. Am. Med. Assoc. 265, 614-617. I141 Viberti, G.C. (1990) New insights into the genesis of diabetic kidney disease in insulin-dependent diabetic patients. In: K.G.M.M. Alberti and L.P. Krall (Eds.), The Diabetes Annual 5, Elsevier BV, Amsterdam, pp. 301-311. 1151Ng, L.L., Simmons, D., Frighi, V. et al. (1990) Leucocyte Na+/H+ antiport activity in type I diabetic patients with nephropathy. Diabetologia 33, 371-377. 1161Viberti, G.C., Keen, H. and Wiseman, J. (1987) Raised arterial pressure in patients of proteinuric insulindependent diabetes. Br. Med. J. 295, 515-517. v71 Hida, M., Iida, T., Watanabe, T. et al. (1990) A clinical study in diabetic nephropathy patients at the time of hemodialysis institution. J. Jpn. Sot. Dial. Ther. 23(10), 1131-l 137. 1181Ritz, E., Nowack, R., Fliser, D., Koch, M. and Tschope, W. (1991) Type II diabetes: is the renal risk adequately appreciated? Nephrol. Dial. Transplant. 6, 679-682.
Retrospective analysis of hemodialyzed diabetic patients in Japan Izumi Takei*, Akira Yamauchi, Shinya Nakamoto, Hiromichi Suzuki, Takao Saruta Department
Of
Internal Medicine, Keio University School
of
.&f&c&,
Shinanomachi
35,
Shinjtku-ku,
T&o,
Japn
160
Received9 May 1995;revision received3 August 1995;accepted8 August 1995
Abstract We retrospectively analyzed the courses of 37 non-insulin dependent diabetics (hemodialyzed:HD group) with endstagerenal disease(ESRD), to identify factors predisposing to renal failure. The factors analyzed were: diabetic (nonproliferative and proliferative) retinopathy, family histories of diabetes and hypertension, smoking, dyslipidemia, first examination proteinuria and non-compliance. These factors were statistically compared in 37 NIDDM without renal failure (non-HD group). There were no significant differences in age or duration of diabetes between the two groups. Significant differences (P < 0.001) were, however, recognized in diabetic proliferative retinopathy and hypertension betweenthe two groups. Hypertension was present in 35/36(97.2%) HD patients and in 21/36(58.3%)non-HD patients. A family history of hypertension was recognized in 16/37HD (43.2%) and in 7/33 (21.2%) non-HD (P < 0.05). Differenceswere recognized in HDL-cholesterol, LDL-cholesterol and TG levels (38.2 l 12.5mg/dl and 56.7 f 18.5mg/dl, 140.4 f 57.1 mg/dl and 115.6 f 33.6 mg/dl, 169.9 f 89.4 mg/dl and 115.7 f 75.1 mg/dl, in HD and non-HD, respectively, P c 0.05). First visit proteinuria was found in all HD patients, and in 6/34 (17.6%) non-HD. The difference in previous treatment refusal, for 7 or more years, was significant with 23/36 (58.9%) HD patients and only l/25 (4.0%) non-HD patients (P < 0.001) having a history of prolonged non-compliance with diabetic treatment. Diabetic retinopathy, non-proliferative and proliferative, hypertension and a family history of hypertension, elevated triglyceride and LDL-cholesterol, low HDL-cholesterol, first visit proteinuria, and prolonged non-compliance correlated with progression to ESRD. We advocate expanding diabetic education to include prevention of complications such as diabetic nephropathy. Keywords: Non-insulin dependent diabetes; End stage renal disease; Hemodialyzed; First visit proteinuria; Noncompliance
1. Introduction
The incidence of renal failure, due to diabetic nephropathy, has increased dramatically in recent l Corresponding author, Tel.: +81 3 33531211 Ext. 2383; fax: +81 3 33592745.
decades. In Japan, the majority of end-stage renal
disease(ESRD) patients are managed with hemodialysis (HD). In addition to the considerable financial burden on medical facilities, HD is associatedwith increased morbidity and mortality in diabetic patients [l-3]. Although
0 1995ElsevierScienceIreland Ltd. All rights reserved 0168~8227/95RO9.50 SSDI 0168-8227(95)01130-6
insulin
dependent diabetics are at
174
I. Takei et al. /Diabetes
Research and Clinical Practice 29 (1995) 173-I 77
higher risk of developing nephropathy and of progressing to ESRD, diabetic nephropathy is also a common finding in those with non-insulin dependent diabetes mellitus (NIDDM) [4]. Reported rates of ESRD in NIDDM patients with nephropathy range from 3- 10% [5-71. As NIDDM is the more common type of diabetes, particularly in Japan, these patients represent the majority of the HD population. It is widely recognized that appropriate medical therapy, especially tight glycemic control, can prevent or delay the development of diabetic nephropathy. Even in those with significant proteinuria (albumin excretion rate > 250 &min), which may occur several years after the diagnosis of NIDDM, loss of renal function can be decelerated with aggressive treatment [8]. Diabetic nephropathy, one of the most serious complications of diabetes, is clinically silent for lo- 15 years. The slow development and long latency period of diabetic nephropathy, while providing ample opportunity for medical intervention, are also major sourcesof patient noncompliance. The problem of noncompliance is one of particular significance in Japan. Only a few decades ago, diabetes was an uncommon diseasein Japan and of little concern to the averageperson. Since the end of the second world war the prevalence of diabetes, especially NIDDM, has increased dramatically. Most Japanesediabetologists attribute this remarkable rise to life-style, particularly dietary, changes. The increased prevalence of NIDDM in our society has not, however, led to increased public awarenessof its complications. The need for prevention of diabetic nephropathy, as well as other serious complications such as retinopathy and neuropathy, is clear. Thus, with the aim of identifying risk factors for the development of diabetic nephropathy, we retrospectively analyzed 37 NIDDM patients on HD. Poor glycemic control, hypertension, a family history of hypertension, smoking, dyslipidemia, and microalbuminuria, as well as the presenceand severity of other diabetic complications, have all been reported to be associated with the development of diabetic nephropathy [2,9,10]. Several of these factors, believed to influence the progression of diabetic nephropathy, were examined in the present study.
2. Subjects and methods The study group consisted of 37 NIDDM patients who had undergone HD at Keio University Hospital within the 5 year period from 1988 through 1992.There were 24 males, 13 femalesand the average age was 62.0 f 10.0 years. The average duration of diabetes before the induction of HD was 17.4 f 7.0 years. The average HbA,c percentagein the 6 months before the induction of HD was 7.0 i 0.9. The control group was made up of 37 NIDDM patients who had never been on HD. There were 19 males and 18 females, with an average age of 65.5 f 11.6years, all of whom had been followed at our out-patient clinic for more than 10 years. All were being treated with insulin or oral hypoglycemic agents. The non-HD and HD groups were well matched in terms of diseaseduration, the average duration of diabetes in the nonHD group being 18.7 f 6.9 years. The average
Table 1 Comparison between the HD and the non-HD groups HD Number of subjects (male/female) Age (mean f SD.) Disease duration (years) HbA,c (%) Complications Proliferative retinopathy Non-proliferative retinopathy No retinopathy Hypertension Family history of hypertension Family history of diabetes mellitus Smoking
non-HD
&3) &8) 62.0 f 10.0 65.5 zt 11.6 17.4 f 7.0 18.7 f 6.9 7.0 l 0.9 8.6 zk l.7* 30136 (83.3%) 6136 (16.7%) O/36 We/,) 35136 (97.2%)
6136 (16.7%)’ 16136 (44.4%)** 14136 (38.9%)* 21136 (58.3%)*
16137 (43.2%) 17136 (47.2%) 15/3l (48.4%)
7133 (21.20/o)*** 21133 (63.6%) lU33 (36.4%)
*P c 0.001; **p c 0.01; l **P c 0.05.
1. Takei et al. /Diabetes
Research and Clinical Practice 29 (1995) 173-177
HbAic percentage in the 6 months prior to this study was 8.6 f 1.7. We periodically measured average lipid metabolic characteristics, specifically total cholesterol (TC), HDL-cholesterol (HDL-C), LDL-cholester01 (LDL-C), and triglycerides (TG), in the HD-group for the 6 month period before the induction of HD and also made thesedeterminations in the non-HD group 6 months prior to starting the study. The results of the two are compared in Table 1. Thirty-six subjects from each group were examined for the existence of diabetic retinopathy, hypertension, family histories of hypertension and diabetes,and smoking, The prevalence of first visit proteinuria, the time course of renal function, and the frequency and duration of periods of noncompliance with treatment were also assessed. The statistical differences were assessed by Fisher’s exact test and non-parametric analysis (Mann Whitney U-test). The data are expressedas means f S.D. A value of P < 0.05 was considered to be statistically significant. 3. Results
Significant differences (P < 0.01) between the HD and non-HD groups were recognized in diabetic retinopathy, diabetic proliferative retinopathy and hypertension. Non-proliferative diabetic retinopathy was identified in 6 out of 36 HD patients (16.7%) and in 16 out of 36 non-HD patients (44.4%), who underwent ophthalmological examination. A significant difference was seen in the prevalence of proliferative diabetic retinopathy, which affected 30 HD patients (83.3%) but
175
only 6 non-HD patients (16.7%) (P < 0.001). Fourteen non-HD patients had no signs of diabetic retinopathy (38.9%). Hypertension was a common finding in both groups, affecting 35 of 36 HD (97.2%) and 21 out of 36 non-HD (58.3%) patients (P < 0.001 ) (Table 1). Detailed family histories were obtained from all of the HD study participants and from 33 non-HD subjects. It was found that 17 of the 36 HD patients (42.2%) and 21 of the 33 non-HD patients (63.3%) had a family history of diabetes, but the difference did not reach statistical significance. A marked difference was, however, recognized for hypertension with 16 of 37 (43.2%) HD but only 7 of 33 (21.2%) non-HD patients reporting a family history of high blood pressure (P < 0.05). Numbers of smokers, 15 of 31 HD and 12 of 33 non-HD patients, were similar in the two groups (Table 1). However no attempts were made to distinguish among light, moderate and heavy smokers or among smoking durations. Those who had quit smoking in the past were considered nonsmokers at the time of the study. Total cholesterol, HDL-cholesterol and serum triglyceride (TG) levels were determined as parameters of lipid metabolism (Table 2). Significant differences were observed in HDL-cholesterol, LDLcholesterol and TG levels. The HDLcholesterol levels were 38.2 f 12.5 and 56.7 f 18.5 (P < O.OOl), LDL-cholesterol levels 140.4 f 57.1 and 115.6 f 33.6 (P < 0.01) and the
I ,CRN
Table 2 Lipid metabolic parameters HD TC (mg/dl) HDL-C (mg/dl) LDLC(mg/dl) TG (mg/dl)
210.7 38.2 140.4 169.9
non-HD zt zt f *
61.6 12.5 57.1 89.4
195.5 56.7 115.6 115.7
f f f f
O!
43.8 18.5* 33.6*** 75.1**
(meanf
S.D.) l P < 0.001; l *P < 0.01; l **p < 0.05.
Fig. 1. Tie
course of renal function (lkxum
creatinine level).
176
1. Takei et al. / Diabeles Research and Chtical Pracfice 29 (1995) /73- 177
Table 3 First visit proteinuria and non-compliance
Proteinuria Non-compliance (7- 11 years without treatment)
HD
non-HD
36/36 (100%) 22/36 (61.1%)
6134(17.6%)* l/25 (4.00/O)*
‘P c 0.001.
serum TG levels 169.9 f 89.4 and 115.7 * 75.1 (P < O.Ol), in the HD and non-HD groups, respectively. The average total cholesterol level 6 months prior to dialysis was higher in the HD group (210.7 f 61.6 as compared to 195.5 f 43.8 in the non-HD group) but the difference was not statistically significant (Table 2). Review of the medical records revealed that proteinuria, of some degreehad been present in all 36 HD patients tested at the initial visit to our outpatient clinic. In contrast, only 6 of the 34 non-HD patients tested had presented with proteinuria (P < 0.001). The reciprocal of the serum creatinine level was used to assessthe time course of renal function. As shown in Fig. 1, over a follow-up period of 70 months, this index fell sharply in the HD group. Among the non-HD patients who had initially presented with proteinuria, there was a slight diminution which correlated with the duration of diabetes. The l/Cm value remained stable in the non-HD patients who did not have proteinuria. The most striking difference identified in this study was in the numbers of patients who had previously refused treatment, for periods ranging 7-l 1 years. An extended period of non-compliance with diabetic therapy and management was reported by all 36 (100%) HD but only 1 of the 25 (4.0%) non-HD patients (P < 0.001 ) (Table 3). 4. Discussion
We conducted a retrospective analysis of HD and non-HD NIDDM patients, for the purpose of identifying risk factors for the development of diabetic nephropathy. As expected, microvascular
complications were present in the majority of those who had been diabetic for more than 10 years. However, the prevalences of both simple and proliferative retinopathy were significantly higher in HD than in non-HD patients. The HD group was also found to have markedly higher rates of hypertension and a family history of hypertension. All but one of the HD patients was hypertensive, and most were treated with calcium channel blockers. As demonstrated by a number of studies in recent years, disorders of lipid metabolism, predisposing to cardiac death, are common fmdings in diabetics with ESRD due to diabetic nephropathy. Our results are in accordance with previous studies concerning association between dyslipidemia and diabetic nephropathy. Levels of HDL-cholesterol were significantly higher, those of TG significantly lower, in the HD than in the non-HD group. The establishment of lipid parameters as predictors of progression to ESRD awaits further clarification of the possible contributions of lipids to the development of diabetic nephropathy. All of the HD patients had initially presented with proteinuria, while less than one fifth of the non-HD patients had been proteinuric at the first visit to our out-patient clinic. Renal function, as assessed by taking the reciprocal of serum creatinine clearance, declined markedly over the following 70 months in the HD group. In contrast, renal function remained relatively stable in the non-HD group. Among the factors examined in this study, neither smoking nor a family-history of diabetes was found to be significantly related to the development of diabetic nephropathy and progression to ESRD. The role of cigarette smoking in the development of albuminuria, reported by Chase et al., may pertain only to insulin-dependent diabetics [13]. The influence of family history, i.e. a genetic predisposition to the development of diabetic nephropathy, has also been investigated primarily in insulin-dependent diabetics [ 14- 161. The role of these factors, if any, in nephropathy and ESRD in NIDDM await further investigation. The striking correlation between extended periods of non-compliance with treatment and
I. Takei et al. /Diabetes
Research and Clinical Praclice 29 (1995) 173-I 77
progression to ESRD has significant educational, as well as prognostic, significance. In recent years, several investigations have emphasized that NIDDM patients are at risk for developing many of the same renal complications as those with IDDM [4,10,17,18]. The risk of developing diabetic nephropathy, with the possibility of progression to ESRD has not been fully appreciated by either patients or clinicians. Diabetes was a relatively uncommon diseasein Japan until 15-20 years ago. As NIDDM is often asymptomatic for several years, many patients diagnosed in the 1970sand 1980sdid not consider it to be a serious health problem. More than half of the HD patients, in the present study, reported having refused treatment for periods of 7-l 1 years. In contrast, only one of the non-HD patients had done so. NIDDM is not simply a very mild form of IDDM. Newly diagnosed NIDDM patients must be fully appraised as to the nature and long-term significance of their disease. Diabetic education, aimed at preventing or slowing the development of complications, is urgently needed. A full understanding of the risks inherent to non-compliance, especially of poor glycemic control and hypertension, ‘will motivate patients to obtain adequate medical care. The prevalence of NIDDM is still rising in Japan. It is not uncommon for proteinuria to be present at, or shortly after, the time of diagnosis. Treatment must not be delayed in such cases. We therefore recommend that diabetic education, aimed at newly diagnosed NIDDM patients, be instituted at the first clinic visit. References [l]
Ghavamian, M., Gutch, C. and Koop, F. (1982) The sad truth about hemodialysis in diabetic nephropathy. J. Am. Med. Assoc. 222, 1368-1369. [2] Matson, M. and Kjellstand, CM. (1988) Long-term follow-up of 369 diabetic patients undergoing dialysis. Arch. Intern. Med. 148, 600-604. (31 Bell, D.S.H. (1991) Diabetic nephropathy: changing concepts of pathogenesis and treatment. Am. J. Med. Sci. 301, 195-200.
117
predicts clini141 Mogensen, C.E. (1984) Microalbuminuria cal proteinuria and early mortality in maturity-onset diabetes. N. Eng. J. Med. 310, 356-360. 151Schimitz, A. and Vaeth, M. (1988) A major risk factor in non-insulin dependent diabetes. A IO-year follow up study of 503 patients. Diabetic Med. 5, 126-134. 161Reddi, A.S. and Camerini-Davalos, R.A. (1990) Diabetic nephropathy: an update. Arch. Intern. Med. 150, 31-43. W.G. (1991) Diabetic I71 Sirmon, M.D. and Kirkpatrick, nephropathy: new direction in management. Renal Failure 13, 51-59. 181Mogensen, C.E. (1976) Progression of diabetic nephropathy in long-term diabetics and effect of initial antihypertensive treatment. Stand. J. Clin. Lab. Invest. 36, 383-388. 191 Gomi, Y., Suzuki, S., Iino, S. et al. (1992) The possibility of hereditary factors in the susceptibility to diabetic nephropathy in NIDDM. Folia. Endocrinol. 68, 592-599. 1101Ritz, E., Raine, A. and Cordonnie, D. (1994) Hemodialysis in Type I and Type II diabetic patients with end stage renal failure. In: C.E. Mogensen (Ed.), The Kidney and Hypertension in Diabetes Mellitus, Kluwer Academic Publishers, Dordrecht, pp. 459-467. 1111Aurell, M. and Bjorck, S. (1992) Determinants of progressive renal disease in diabetes mellitus. Kidney. lnt. Suppl. 36, S38-S42. 1121Tschope, W., Koch, M., Thomas, B. and Ritz, E. (1993) Serum lipids predict cardiac death in diabetic patients on maintenance hemodialysis. Nephron 64. 354-358. 1131 Chase, S.P., Garg, S.K. and Hamman, R.E. et al. (1991) Cigarette smoking increases the risk of albuminuria among subjects with type I diabetes. J. Am. Med. Assoc. 265, 614-617. I141 Viberti, G.C. (1990) New insights into the genesis of diabetic kidney disease in insulin-dependent diabetic patients. In: K.G.M.M. Alberti and L.P. Krall (Eds.), The Diabetes Annual 5, Elsevier BV, Amsterdam, pp. 301-311. 1151Ng, L.L., Simmons, D., Frighi, V. et al. (1990) Leucocyte Na+/H+ antiport activity in type I diabetic patients with nephropathy. Diabetologia 33, 371-377. 1161Viberti, G.C., Keen, H. and Wiseman, J. (1987) Raised arterial pressure in patients of proteinuric insulindependent diabetes. Br. Med. J. 295, 515-517. v71 Hida, M., Iida, T., Watanabe, T. et al. (1990) A clinical study in diabetic nephropathy patients at the time of hemodialysis institution. J. Jpn. Sot. Dial. Ther. 23(10), 1131-l 137. 1181Ritz, E., Nowack, R., Fliser, D., Koch, M. and Tschope, W. (1991) Type II diabetes: is the renal risk adequately appreciated? Nephrol. Dial. Transplant. 6, 679-682.