Coronary Artery Disease Is the Major Determinant of Excess Mortality in Patients With Insulin-Dependent Diabetes Mellitus and Persistent Proteinuria

Transcription

1 Coronary Artery Disease Is the Major Determinant of Excess Mortality in Patients With Insulin-Dependent Diabetes Mellitus and Persistent Proteinuria James H. Warram,1 Lori MB. Laffel, Om P. Ganda, and A. Richard Christlieb J. H. Warram, L. M. B. Laffel,. P. Ganda, A. P. Christlieb, Research and Clinic Divisions, Joslin Diabetes Center, Boston, MA (J. Am. Soc. Nephrol. 1992; 3:14-11) ABSTRACT The goal of this review was to assess the magnitude of coronary artery disease (CAD) mortality and its determinants in insulin-dependent diabetes mellitus (IDDM) patients with persistent proteinuria. By reanalyzing data from two previously published studies of patients with nephropathy, it was found that these patients had extremely high CAD mortality rates in comparison with IDDM patients without proteinuria, but only after the age of 35 yr. In addition, the risk of CAD death was associated with high serum cholesterol levels but was unrelated to systemic blood piessure, smoking habits, and obesity. Further studies of the determinants of CAD in patients with IDDM and proteinuria are urgently needed. Except for efforts to lower serum cholesterol, it is not known whether any other measure can be undertaken to reduce the extremely high mortality due to CAD that afflicts IDDM patients with persistent proteinuria, in particular those patients whose renal failure might have been successfully postponed by antihypertensive therapy. Key Words: Insulin-dependent diabetes, persistent proteinuria. coronary artery disease, hypertension, serum cholesterol T he onset of persistent protcinuria is the first clinical evidence of progressive renal injury in juvenile-onset Insulin-dependent diabetes mebbitus (IDDM) (1,2). In addition to heralding renal failure, protcinuria is also a marker of excess risk of cardiovascular disease, particularly affecting the coronary arteries (coronary artery disease [CAD]) (3,4)., Correspondence to Dr. J. H. Warram, Epidemiology and Genetics, Joslin Diabetes Center, One Joslin Place, Boston, MA /34-S14$3./ Journal of the American Society of Nephrology Copyright C 1992 by the American Society of Nephrology There Is consensus that efforts to control blood pressure preserve kidney function (5,6). A very different situation prevails In the case of cardiovascular disease among patients with diabetic ncphropathy. There has been no clinical trial of this Issue In patients with!ddm. In nondiabetics, the clinical trials of antlhypertensive therapy have demonstrated only modest reduction in CAD mortality (7). Thus, this indirect evidence does not give assurance that prolongation of kidney function by lowering blood pressure in!ddm will significantly reduce the threat of death due to CAD. The goal of this review is to examine available data in order to assess the magnitude of CAD mortality and its determinants in IDDM patients with ncphropathy. In particular, we will examine the role of hypertension and other cardiovascular risk factors, such as age, gender, smoking, and lipids, in order to design interventions that might reduce the risk of death due to CAD. DESCRIPTION OF STUDIES FROM WHICH DATA WERE REANALYZED There arc two studies from which we reanalyzed data on the natural history of persistent proteinurla In!DDM. The first is a study of a cohort of 292 patients wlthjuvcnlbc-onsct type I diabetes who were monitored from the diagnosis of diabetes in 1939, or until (2 to 4 yr of follow-up) (2,4). Persistent proteinuria (two out of three consecutive urinalyses showed a protein concentration of 3 mg/dl or more) developed in 67 of them, and by the end of follow-up observation, 46 had died, 8 were surviving with a kidney transplant or dialysis, and 1 3 remained in the stage of persistent proteinuria. Although the study group is not large, it provides a complete view of the natural history of diabetic nephropathy, from diagnosis until death. The second study has complementary strengths. The Diabetic Retinopathy Study (DRS) was a multicenter, randomized, clinical trial of laser therapy for the prevention of blindness due to proliferative diabetic retinopathy (8,9). Fifteen clinical centers enrolled 1,758 patients in the trial and monitored them until 1979, an average of 5 yr. All patients had a confirmed diagnosis of diabetes and severe diabetic retinopathy at the beginning of follow-up. Data recorded at the annual follow-up examinations in- S I 4 Volume 3 Supplement I 1992

2 C Warram et al c!udcd height, weight, blood pressure, treatment for blood pressure, urine protein, and serum creatinine. Vital status was monitored three times a year, and information regarding cause of death was collected and reviewed by a mortality committee. In this study group, there were 443 patients who fulfilled the fobbowing selection criteria: insulin-treated diabetes diagnosed before age 3, persistent proteinuria (defined on the basis of the baseline and first annual follow-up examinations [9]), and serum creatinine below 2.5 mg/dl. In contrast to the cohort study, follow-up observation is short, a maximum of 7 yr for each person, but there arc annual measurements of a fixed set of clinical variables. IDDM PATIENTS WITH DIABETIC NEPHROPATHY HAVE POOR SURVIVAL The natural history study of nephropathy in juvenile-onset, type I diabetes conducted at the Joslin Clinic Included 67 patients (4 men, 27 women) who were monitored from the onset of persistent proteinuria (2,4). The average age at the time of the diagnosis of IDDM was 1 2 yr (all were under the age of 2 1). The average age at the time of the diagnosis of persistent protelnuria was 28 yr, and In calendar time, these onsets were equably distributed over three decades, the 195s, 196s, and 197s. The cumulative survival of this cohort after the diagnosis of persistent protcinurla was calculated separately according to the decade of the diagnosis of diabetic nephropathy (Figure 1). The natural history of proteinurla was quite prolonged, some cases surviving more than 2 yr. Median survival was around 1 yr for those whose nephropathy was diagnosed in the 1 95s and was exactly the same in > (1) C a U a s 196s 195s Years of Survival Figure 1. Cumulative survival from the onset of persistent proteinuria in patients with juvenile-onset IDDM according to decade of onset of nephropathy (see Text). There were 17 patients in the 195s cohort, 27 in the 196s cohort, and 23 in the 197s cohort. A log-rank test of equality of the survival curves gave a x2 of.29; P>.86. each successive decade, even though the management of nephropathy was evolving over time. For example, most patients In the group diagnosed in the 1 97s were treated with antihypertensive agents to control blood pressure, whereas the 1 95s cohort would not have received diuretics until late In the disease course to control edema. The back of improvement In survival between the cohorts from the 1 95s and 1 97s stands In marked contrast to a report of improved survival in similar cohorts followed at Steno Memorial Hospital in Denmark (1 ). Part of the reason for the difference Is the uncharacteristically poor survival of the early Danish cohort. Survival was only 46% after 8 yr of proteinuria In the Danish cohort assembled between 1957 and 1973, whereas 8-yr survival was 69% In the Joshin patients assembled during the 1 95s, as we!! as In those assembled in the 1 96s and 1 97s. In the Danish cohort assembled between and 1983, survival was much improved. 87% at 8 yr. Whether there has been similar improvement during the 1 98s in the survival of proteinuric patients at the Joslin Clinic remains to be studied. The Danish authors attributed the improved survival of their patients to better control of hypertension with various antihypertensive agents in the later time period. Hypertension was defined as diastolic blood pressure >95 mm Hg. As noted above, the treatment of hypertension came into vogue at the JoshIn Clinic in the 1 97s, yet survival was 69% at 8 yr, just as It had been during earlier decades. INCREASING AGE IS THE MAJOR DETERMINANT OF CAD DEATH Table 1 shows the mortality rates in the two studies according to attained age. The mortality rate in patients with proteinuria in the cohort study was simibar to the mortality rate In!DDM patients with proteinurla in the DRS, and these rates were several times higher than among!ddm patients without proteinuria. The mortality rate among patients with persistent proteinuria Is profoundly affected by age (Table 1). Before age 35 (7 yr after the median age at onset of protcinurla), It was 44 per 1 per year, whereas after age 35, It nearby tripled to 1 16 per 1 per year. Deaths among young patients were predominatehy renal deaths, whereas after age 35. the CAD mortality rate equaled the renal mortality rate (Table 2). Note that two thirds of the 2-yr natural history of ncphropathy occurs after age 35. If the progression of nephropathy is slowed by various interventions, the prolonged survival will extend the years that these patients are exposed to a very high risk of CAD mortality. Thus, it is Important to Identify and modify the determinants of CAD mortality in this population Journal of the American Society of Nephrology SI 5

3 Diabetic Nephropathy and CAD Mortality TABLE I. Comparison of the mortality rate (per 1,) in the Cohort Study and the DRS in order to prolong life rather than merely trade one cause of death for another. Study Group Mortality Rate (per 1, person yr) Age Cohort Study#{176} Without proteinuria With proteinuria DRS with proteinuri& All Ages A total of 282 patients in the cohort study survived and were monitored beyond age 25. giving a total of 3,983 person yr of observation. Persistent proteinuria developed in 67 of them. and they were observed for a total of 61 1 person yr after the onset of proteinuria. leaving a total of person yr of observation without proteinuria ( person yr). b There were 443 patients with IDDM and persistent proteinuria. and they were observed for a total of I.69 person yr. TABLE 2. Cause-specific mortality rates (per 1,) in a cohort of patients with IDDM and persistent proteinuria according to age Cause of Death Ag e (yr) All Ages CAD Renal Other 3 2 Total Number of Deaths Number of Person Yr COMPARISON OF IDDM PATIENTS WITH NEPH- ROPATHY WHO DIED AS A RESULT OF CAD AND THOSE WHO SURVIVED To examine the determinants of CAD mortality in patients with early nephropathy (before the stage of progressive renal failure), we restricted the study to patients with normal serum creatlninc (<1.5 mg/dl) at the beginning of follow-up and then subdivided them according to whether their serum creatinine had risen to 2.5 mg/dl by the time of death or the end of follow-up (FAST PROGRESSORS) or whether it remained below 2.5 mg/dl (SLOW PROGRES- SORS). Regardless of the rate of progression of renal disease, CAD was the major cause of death (Table 3). We selected the CAD deaths, 14 among the FAST PROGRESSORS and 24 among the SLOW PROGRES- SORS, as CASES. Then, from among the patients who survived to the end of follow-up, we selected as CONTROLS those who were monitored for at beast 4 yr after randomization into the DRS (22 FAST PRO- GRESSORS and 129 SLOW PROGRESSORS). The CONTROLS and the CASES (CAD deaths) were the same age, had similar body weight, and a similar percentage from each group were smokers (Table 4). There were fewer males among CASES than CON- TROLS, but the difference was not statistically significant. Furthermore, it is important to note that because they were matched according to renal status at the beginning and end of follow-up, they had equal serum creatinine levels at the beginning of the study and the course of serum creatinine from baseline to the end of the study was also similar (data not shown). TABLE 3. Distribution of the study group according to serum creatinine at the end of the study and according to outcome at the end of follow-up Final serum creatinine Outcome (<2.5 mg/dl) (2.5 mg/dl) SLOW PROGRESSORS FAST PROGRESSORS (N) (N) Survived and Monitored <4yrs yrs Death in yr 3-7 by Cause CAD Renal 5 5 Other 2 Total number SI 6 Volume 3. Supplement I

4 Warram et al TABLE 4. Characteristics of the CAD deaths and survivors at entry into the study according to whether they had rapidly or slowly progressive renal disease Characteristic/Progression Survivors CAD Deaths Age at Entry (yr) Fast 29±6 29±7 Slow 31±1 32±11 Duration of IDDM (yr) Fast 17±3 15±4 Slow 19±7 19±7 Age at Diagnosis of IDDM (yr) Fast 12±6 14±8 Slow 12±7 13±8 Percent Ideal Body Weight Fast 16 ± ± 33 Slow 15 ± ± 29 Male (%) Fast Slow Cigarette Smoker (%) Fast 6 36 Slow 57 6 TABLE 5. Blood pressures at beginning of the study and at the last examination before death or end of follow-up Progression Survivors CAD Deaths Baseline Blood Pressure Systolic Fast 146±2 152±3 Slow 139 ± ± 2 Diastolic Fast 89±12 93±15 Slow 84±1 84±9 Mean Follow-Up Blood Pressureb Systolic Fast 157 ± ± 14 Slow 138±19 145± 19 Diastolic Fast 96±1 98±9 Slow 85±1 88±1 Last Blood Pressure Systolic Fast 165 ± ± 21 Slow 142±21 142± 28 Diastolic Fast 95±13 99±1 Slow 86±12 88±15 ELEVATED BLOOD PRESSURE DOES NOT DIS- CRIMINATE THE RISK OF CAD IN PATIENTS WITH NEPHROPATHY The effect of blood pressure on the risk of CAD was examined by comparing blood pressure bevels in CASES and CONTROLS at several points in time (Table 5), separately among FAST and SLOW PRO- GRESSORS. At baseline, systolic and diastolic blood pressures were similar In CAD deaths and survivors, although both were higher In the FAST PROGRES- SORS than in the SLOW PROGRESSORS. Similarly, the average blood pressures during follow-up and at the last visit before death or the end of follow-up were higher in the FAST PROGRESSORS. There was some increase In blood pressure during follow-up, more so in FAST PROGRESSORS than in SLOW PROGRESSORS, but this trend was no greater among these CAD deaths than in survivors. Thus, there was no period during which the CAD deaths had higher blood pressures than the survivors. The back of a difference in blood pressure between CASES and CONTROLS was not due to a greater use of antihypertensive treatment in CASES. From the start, the majority of patients in each group had hypertension (systolic blood pressure 1 4 mm Hg, diastolic blood pressure 9 mm Hg, or antihypertensive treatment): 65% of CASES and 7% of CON- TROLS among SLOW PROGRESSORS and 82% of CASES and 93% of CONTROLS among FAST PRO- GRESSORS. Yet the corresponding proportions Values are mean ± SD. b Average of blood pressure measurements on second through fourth follow-up visits. treated with antihypertensives during follow-up were 24 and 25% among SLOW PROGRESSORS and 45 and 54% among FAST PROGRESSORS. Treatment for hypertension was not a consistent policy in the decade (between and 1 979) when the DRS was being conducted, but those who were treated did not have a bower risk of death due to CAD than those who were not. RISK OF CAD IN IDDM PATIENTS WITH NEPH- ROPATHY IS RELATED TO LIPID ABNORMALITIES In contrast to the absence of an association between CAD mortality and blood pressure, there was a significant association between CAD mortality and total serum cholesterol. Serum cholesterol at the baseline examination was slightly higher among FAST PROGRESSORS than among the SLOW PRO- GRESSORS but was substantially higher in CAD deaths than In survivors, regardless of renal status: 27 ± 62 as compared with 237 ± 45 among FAST PROGRESSORS and 264 ± 59 as compared with 22 ± 43 among SLOW PROGRESSORS. The association of CAD death with serum cho!csterol can be assessed more graphically by comparing the distributions of serum cholesterol rather than the Journal of the American Society of Nephrology 517

5 Diabetic Nephropathy and CAD Mortality,... - c it Cl) #{149} V Serum D<24o Cholesterol mg/dl 1>24 mg/dl Slow Progressors Fast Progressors Figure 2. Relative odds of death due to CAD associated with serum cholesterol (total) above 24 mg/dl at entry to the DRS according to the rate of progression of renal disease (see Text for definitions). Confidence interval for relative odds among SLOW PROGRESSORS is 1.6 to 12; among FAST PROGRESSORS, it is.6 to 9.2. means. Less than 3% of the control group had a cholesterol above 24 mg/dl. whereas cholesterol above this value was much more common in each of the groups of CAD deaths. This can be summarized as the relative odds or relative risk of coronary death according to serum cholesterol (Figure 1). Among SLOW PROGRESSORS, a cholesterol above 24 mg/ dl was associated with a 4.4-fold risk of coronary death relative to a lower cholesterol (P <.5). Among FAST PROGRESSORS, the relative risk was 2.3, which was not statistically significant with the small sample size. IMPLICATION FOR FUTURE RESEARCH AND CLINICAL MANAGEMENT!DDM patients with persistent proteinurla have very high mortality in comparison with IDDM patients without this complication (Tables 1 and 2). Before age 35, the excess mortality is primarily due to renal failure, whereas after age 35, the mortality rate due to CAD equals or exceeds that related directly to the underlying kidney disease, each being responsib!c for the deaths of about 5% of the population per year. The mechanisms responsible for the emergence of CAD after age 35 as the major complication in IDDM patients with nephropathy is unclear. One possibihity is that the metabolic and coagulation abnormabities that precede the development of persistent proteinurla and that are magnified during that stage may accelerate atherosclerosis and manifestation of CAD clinically only if the atherosclerotic lesions had been initiated (4, 1 1). This hypothesis Is consistent with observations that showed that patients with diabetic nephropathy had a very low risk of CAD in populations in which CAD is infrequent In the background populations, for example in Japanese and Pima Indian patients (12,13). The importance of age as a determinant of CAD in white patients with diabetic nephropathy must be seen in the context of studies that have shown that the loss of renal function during the stage of persistent proteinuria could be retarded by controlling dcvated systemic blood pressure (5,6). As a result of such programs, IDDM patients with ncphropathy will become older and might avoid deaths due to renal failure but die as a result of CAD. Using the follow-up data carefully collected for the DRS, we have sought to identify determinants of mortality due to CAD in IDDM patients with persistent proteinuria. The findings arc somehow disappointing. We found a back of association between most of the conventional risk factors and deaths due to CAD. Patients with persistent protdinuria who died as a result of CAD were not different from similar patients, matched for rena! status, who survived 4 to 7 yr with regard to age, gender, smoking habits, percent ideal body weight, or bevels of blood pressure. Regarding the last of these, arterial blood pressure at the baseline examination was positively associated with the rate of progression of renal disease but seemed to have no bearing on whether FAST PRO- GRESSORS or SLOW PROGRESSORS died of coronary disease. Moreover, blood pressure increased in all groups as renal disease worsened but It was similar in those who died of CAD and those who survived. This lack of effect of blood pressure elevation, whether antecedent or concurrent, on the risk of CAD is difficult to explain. First, it must be viewed in the context that patients who develop persistent protdinuria have a predisposition to essential hypertension and, frequently, have a history of coronary disease in relatives (1 4, 1 5). As expected, therefore, most patients in this study were hypertensive and, presumably, were exposed to higher than average blood pressures for many years before the development of proteinuria. The lack of association between hypertension and CAD in this study. therefore, may be because virtually all had this predisposition, so factors other than hypertension determined which ones died of CAD during the short follow-up interval. The second consideration is the stage in the natural history of atherogenesis at which hypertension has its effect. If it is early, perhaps on the initial lesions, modification of systemic blood pressure may not reduce mortality due to CAD unless treatment is begun early. In fact, blood pressure reduction late in the process may compromise myocardial perfusion if sig- SI 8 Volume 3. Supplement I. 1992

6 Warram et al nificant stenosis is present ( 1 6, 1 7) or autonomic neuropathy has developed (18,19). In contrast to the lack of association between dcvated blood pressure and mortality due CAD, we found that elevated serum cholesterol was strongly associated with death due to CAD. To interpret this finding, we should consider two possibilities. First, elevated cholesterol may contribute to the development of atherosclerosis bong before the clinical manifestation of persistent proteinuria. In that case, the association found by us merely indicates that patients with hyperchobesterolemia have rather advanced atherosclerotic lesions from the start of diabetic ncphropathy. The other possibility is that the development of persistent proteinuria itself causes significant lipid abnormalities that accelerate the progression of advanced atherosclerotic lesions. Both possibilities may operate in the same patient. Regardless of which is true, hypercholesterobemla seems to Influence the progression of bate stages of atherogenesis, a process that might be diminished by efforts to reduce lipid abnormalities in persistent protelnurla. To distinguish among these possibilities, one must conduct a much longer follow-up study, perhaps starting with normoabbuminuric IDDM patients. Meanwhile, the possibility of conducting a clinical trial should be considered. IDDM patients with persistent microalbuminuria/proteinuria should be randomized to subgroups that will be treated with varbus lipid-lowering programs to determine whether the risk of CAD is modified. Finally, the association of baseline cholesterol with death due to CAD in IDDM patients with ncphropathy does not account for all of the excess of CAD in this population. Other determinants must be sought, including other lipid abnormalities such as hypcrtrig!yccridemia or elevated bevels of apobipoprotcin (a) (2.21). Cardiovascular autonomic neuropathy, which is very common In patients with diabetic ncphropathy and proliferative diabetic retinopathy, should also be considered as another factor that may contribute to high mortality due to CAD (18,19,22). ACKNOWLEDGMENTS This research was partially funded by Diabetes Research & Education Foundation Grant no. C5552/II-9 1. REFERENCES 1. Andersen AR, Christiensen JS, Andersen JK, Kreiner 5, Deckert T: Diabetic ncphropathy in type I (Insulin-dependent) diabetes: An epidemiobogic study. Diabetobogia 1 983;25: Krolewski AS, Warram JH, Christlieb AR, Busick EJ, Kahn CR: The changing natural history of nephropathy In type I diabetes. Am J Med 1 985;78: Deckert T, Poulsen JE, Larsen M: Prognosis of diabetics with diabetes onset before the age of thirty-one. I. Survival, causes of deaths and comphications. Diabetobogia 1 978; 14: Krolewski AS, Kosmski EJ, Warram JH, et at.: Magnitude and determinants of coronary artery disease In juvenile-onset, insulin-dependent diabetes melhitus. Am J Cardiob 1987;59: Parving HH, Andersen AR, Smidt UM, Hommel E, Mathiesen ER, Svendsen PA: Effect of antihypertensive treatment on kidney function in diabetic nephropathy. BMJ 1 987;294: Bjorck 5, Mulec H, Johnsen SA, Norden G, Am-eli M: Renal protective effect of enabaprib in diabetic nephropathy. BMJ 1992;34: Hebert PR, Manson JE, Hennekens CH: Pharmacologic therapy of mild to moderate hypertenslon. Possible generahizabihity to diabetics. J Am Soc Nephrol 1992, in press. 8. Diabetic Retinopathy Study Group: Diabetic Retinopathy Study: Report number 6: Design, methods, and baseline results. Invest Ophthalmob Vis Sd 1981;21: Warram JH, Laffel LMB, Valsania P, Christlieb AR, Krolewski AS: Excess mortality associated with diuretics therapy In diabetes mellitus. Arch Intern Med 1991;151: Mathiesen ER, Borch-Johnsen K, Jensen DV, Deckert T: Improved survival In patients with diabetic nephropathy. Diabetobogla 1 989;32: Krolewski AS, Warram JH, Valsania P, Martin BC, Laud LMB, Christlieb AR: Evolving natural history of coronary artery disease In diabetes melbitus. Am J Mcd 1991 ;9(suppb. 2A): Kikkawa R, Araki 5, Haneda M, Kajiwara N, Hidaka H, Shigeta Y: Hypertension and the devebopmcnt of complications in patients with noninsuhin dependent diabetes melbitus in Japan. J Am Soc Ncphrol 1992, In press Nelson RG, Sievers ML, Knowler WC, et at.: Low incidence of fatal coronary heart disease In Pima Indians despite high prevalence of noninsulin dependent diabetes. Circulation 199; 81: Krolewski AS, Canessa M, Warram JH, et at.: Predisposition to hypertension and susceptibil- Ity to renal disease In insulin-dependent diabetes melhitus. N Engl J Mcd 1988;318: Earle K, Walker J, Hill C, Viberti GC: Familial clustering of cardiovascular disease in patients with insulin-dependent diabetes and nephropathy. N Engb J Med 1992;326: Cruickshank JM: Coronary flow reserve and the J-curve relation between diastolic blood pressure and myocardiab infarction. BMJ 1988;297: Alderman MH, Ooi WL, Madhavan S, Cohen H: Treatment-induced blood pressure reduction and the risk of myocardiab infarction. JAMA 1 989;262: Ewing DJ, Campbell 1W, Clarke BF: The natural history of diabetic autonomic neuropathy. Q J Med 1 98; 193: Page MM, Watkins PJ: Cardiorespiratory arrest in diabetic autonomic neuropathy. Lancet 1978; 1: Journal of the American Society of Nephrology SI 9

7 Diabetic Nephropathy and CAD Mortality 2. Jenkins AJ, Steele JS, Janus ED, Best JD: Increased plasma apobipoprotcin(a) bevels in IDDM patients with microabbuminuria. Diabetes 1991 ;4: Kapeirud H, Bangstad HJ, DaM-Jorgensen K, Berg K, Hanssen KF: Serum Lp(a) hipoprotein concentrations In insulin dependent diabetic patients with microalbuminuria. BMJ 1991 ;33: Krolewski AS, Barzilay J, Warram JH, Martin BC, Pfeifer M, Rand LI: Risk of early-onset prohiferative retinopathy In IDDM Is closely rebated to cardiovascular autonomic neuropathy. Diabetes 1992;41 : Silo Volume 3. Supplement I. 1992