Clinical Investigation and Reports

Transcription

1 Clinical Investigation and Reports Cardiovascular Events and Their Reduction With Pravastatin in Diabetic and Glucose-Intolerant Myocardial Infarction Survivors With Average Cholesterol Levels Subgroup Analyses in the Cholesterol And Recurrent Events (CARE) Trial Ronald B. Goldberg, MD; Margot J. Mellies, MD; Frank M. Sacks, MD; Lemuel A. Moyé, MD, PhD; Barbara V. Howard, PhD; William James Howard, MD; Barry R. Davis, MD, PhD; Thomas G. Cole, PhD; Marc A. Pfeffer, MD, PhD; Eugene Braunwald, MD; for the CARE Investigators Background Although diabetes is a major risk factor for coronary heart disease (CHD), little information is available on the effects of lipid lowering in diabetic patients. We determined whether lipid-lowering treatment with pravastatin prevents recurrent cardiovascular events in diabetic patients with CHD and average cholesterol levels. Methods and Results The Cholesterol And Recurrent Events (CARE) trial, a 5-year trial that compared the effect of pravastatin and placebo, included 586 patients (14.1%) with clinical diagnoses of diabetes. The participants with diabetes were older, more obese, and more hypertensive. The mean baseline lipid concentrations in the group with diabetes 136 mg/dl LDL cholesterol, 38 mg/dl HDL cholesterol, and 164 mg/dl triglycerides were similar to those in the nondiabetic group. LDL cholesterol reduction by pravastatin was similar (27% and 28%) in the diabetic and nondiabetic groups, respectively. In the placebo group, the diabetic patients suffered more recurrent coronary events (CHD death, nonfatal myocardial infarction [MI], CABG, and PTCA) than did the nondiabetic patients (37% versus 25%). Pravastatin treatment reduced the absolute risk of coronary events for the diabetic and nondiabetic patients by 8.1% and 5.2% and the relative risk by 25% (P 0.05) and 23% (P 0.001), respectively. Pravastatin reduced the relative risk for revascularization procedures by 32% (P 0.04) in the diabetic patients. In the 3553 patients who were not diagnosed as diabetic, 342 had impaired fasting glucose at entry defined by the American Diabetes Association as 110 to 125 mg/dl. These nondiabetic patients with impaired fasting glucose had a higher rate of recurrent coronary events than those with normal fasting glucose (eg, 13% versus 10% for nonfatal MI). Recurrence rates tended to be lower in the pravastatin compared with placebo group (eg, 50%, P 0.05 for nonfatal MI). Conclusions Diabetic patients and nondiabetic patients with impaired fasting glucose are at high risk of recurrent coronary events that can be substantially reduced by pravastatin treatment. (Circulation. 1998;98: ) Key Words: diabetes mellitus coronary disease glucose lipoproteins Abnormal glucose tolerance is a major risk factor for coronary heart disease (CHD), increasing its likelihood by a factor of 2 to 4, 1,2 and the prevalence of diabetes is at least 2- to 3-fold higher among patients with established CHD than among the general population. 3 The basis for this wellestablished association between diabetes and CHD is not understood. Undoubtedly, the increased frequency of conventional cardiovascular risk factors such as dyslipidemia and hypertension in diabetes is an important determinant of CHD in these patients. 4,5 Whether hyperglycemia per se contributes to this increased risk is controversial. 6 8 Recently, a large, 22-year prospective study demonstrated that mild hyperglycemia was associated with excess cardiovascular mortality in men. 9 Despite the heightened risk for CHD, patients with known glucose intolerance have largely been excluded from trials of cholesterol-lowering therapy. A recent report from the Scandinavian Simvastatin Survival Study (4S) indicated that in the subgroup of diabetic patients with hypercholesteremia and Received June 22, 1998; revision received August 3, 1998; accepted August 13, From the University of Miami School of Medicine, Miami, Fla (R.B.G.); the University of Texas School of Public Health, Houston, Tex (L.A.M., B.R.D.); Bristol Myers Squibb, Princeton, NJ (M.J.M.); Medlantic Research Institute, Washington, DC (B.V.H.); Washington Hospital Center, Washington, DC (W.J.H.); Washington University, St. Louis, Mo (T.G.C.); and Brigham and Women s Hospital and Harvard Medical School, Boston, Mass (F.M.S., M.A.P., E.B.). Correspondence to Frank M. Sacks, MD, Nutrition Department, Harvard School of Public Health, 665 Huntington Ave, Boston, MA fsacks@ hsph.harvard.edu 1998 American Heart Association, Inc. Circulation is available at

2 2514 Pravastatin, Diabetes, and CHD normal triglycerides with established CHD, lowering of LDL cholesterol (LDL-C) levels with the HMG CoA reductase inhibitor simvastatin was associated with a significant reduction in cardiovascular morbidity and mortality. 10 However, because most patients with diabetes and CHD do not have significant hypercholesterolemia, 7 the relevance of this finding to diabetic patients with CHD and average cholesterol levels is unclear. The Cholesterol And Recurrent Events (CARE) trial recently demonstrated that cardiovascular events are significantly reduced in patients with average cholesterol levels after myocardial infarction (MI) who are treated with the HMG CoA reductase inhibitor pravastatin. 11 The CARE study population contained a relatively large cohort of diabetic patients, thus providing an opportunity to evaluate the effect of pravastatin treatment on the recurrence of CHD in these patients. In addition, the predictive value of fasting glucose levels for recurrent CHD events and the effects of pravastatin treatment in nondiabetic patients with impaired fasting glucose (110 to 125 mg/dl) were examined. Methods Design of the CARE Trial The study design has been described in detail elsewhere. 11 Briefly, men and postmenopausal women between 21 to 75 years of age who had suffered an MI between 3 and 20 months before randomization who had plasma total cholesterol values 240 mg/dl, LDL-C levels between 115 and 174 mg/dl, and triglycerides 350 mg/dl were randomized into pravastatin (40 mg/d) or placebo treatment groups in a double-blind manner. Patients with fasting glucose levels 220 mg/dl, left ventricular ejection fractions of 25%, and symptomatic congestive heart failure were excluded. All participants entered into a dietary program supervised by a registered dietitian using the approach of the National Cholesterol Education Program. 13 A total of 4159 patients were randomized in 80 clinical centers in the United States and Canada: 2081 to pravastatin and 2078 to placebo treatment. A serum glucose measurement was performed at the baseline visit for 4133 of the patients (99.4%) after an overnight fast by the core laboratory by the hexokinase method with an autoanalyzer (Technicon). Plasma lipids, drawn after an overnight fast, were measured 2 to 3 times at baseline and at intervals thereafter at a core laboratory that is certified for lipid measurements by the Centers for Disease Control (Washington University). LDL-C levels were calculated. 14 The primary end point of the trial was the combination of death from CHD plus nonfatal MI, and the average duration of follow-up was 5 years. An expanded end point was defined as the primary end point, bypass surgery, or angioplasty and was used for subgroup analysis. 12 The protocol was approved by the institutional review boards of all participating centers. All patients randomized into CARE were interviewed and asked whether they previously had been informed that they had diabetes or had received medication for diabetes. Positive responders constituted the diabetes group (n 586, 14.1%), and all others in CARE constituted the nondiabetes group (n 3573, 85.9%). Patients who reported an absence of diabetes were further divided into categories established by the American Diabetes Association 15 based on fasting blood glucose concentrations: normal fasting glucose ( 110 mg/dl) and impaired fasting glucose (110 to 125 mg/d). Statistical Analysis All analyses were performed on an intention-to-treat basis, and probability values were 2-sided. The effects of pravastatin compared with placebo on the rate of the primary and expanded end points of the trial were assessed with the use of the log-rank probability values separately for the diabetes and nondiabetes groups and for the groups with normal and impaired fasting glucose. 16 All other hypothesis tests and all reductions in risk were assessed in each of the groups TABLE 1. Demographic and Clinical Characteristics at Baseline Diabetes No Diabetes P Participants 586 (14.1) 3573 (85.9) Male sex 471 (80.4) 3112 (87.1) Age, y 61 (8) 58 (9) White 498 (85.0) 3353 (93.8) No alcohol intake 442 (75.4) 1955 (54.7) Current smoking 95 (16.3) 606 (21.4) 0.02 Family history of heart disease 253 (39.8) 1442 (40.4) NS CABG 152 (25.9) 744 (20.8) PTCA 127 (21.7) 1022 (28.6) Stroke 35 (6.0) 86 (2.4) TIA 16 (2.7) 87 (2.4) NS CHF history 73 (12.5) 223 (6.2) Hypertension history 302 (51.5) 1472 (41.2) Gout 27 (4.6) 149 (4.2) NS Current claudication 80 (13.7) 242 (6.8) Aspirin 458 (78.2) 3015 (84.4) Blockers 229 (39.1) 1439 (40.3) NS Calcium channel blockers 199 (34.0) 1020 (28.5) ACE inhibitors 126 (21.5) 474 (13.3) Diuretics 136 (23.2) 357 (10.0) Digitalis 75 (12.8) 254 (7.1) Nitrates 235 (40.1) 1143 (32.0) TIA indicates transient ischemic attack; CHF, congestive heart failure. Values shown are n (%) except for age, which is mean (SD). with a Cox proportional-hazards model with or without adjustment for differences in age and sex. 17 Relative risk reductions in the diabetes and nondiabetes groups and in the normal and impaired fasting glucose groups were compared by including both groups in the Cox proportional-hazards models and adding an interaction term, eg, diabetes times treatment group. Results The baseline characteristics of CARE participants with and without previously diagnosed diabetes are shown in Tables 1 and 2. A clinical diagnosis of diabetes was reported in 586 patients, ie, 14.1% of the CARE population. Of these, 45.4% were receiving sulfonylurea treatment, and 18.6% were treated with insulin. Compared with patients without diabetes, those with diabetes were older and more frequently obese, smoked less, drank alcohol less, and were more likely to be female or a member of a minority ethnic group. They more frequently had a history of hypertension, intermittent claudication, cerebrovascular accident, and congestive heart failure than those without diagnosed diabetes, and they had higher mean systolic blood pressures and heart rates. Patients with diabetes were also more likely to be receiving digitalis, nitrates, ACE inhibitors, and diuretics. Cholestyramine was used by 3% of the diabetics compared with 6% of the nondiabetics. The group with diabetes had slightly but significantly lower mean LDL-C and HDL-C and higher mean triglyceride levels than the nondiabetic group (Table 2). The baseline mean fasting glucose value in those with

3 Goldberg et al December 8, TABLE 2. Physical and Laboratory Findings Diabetes (n 586) No Diabetes (n 3573) BMI, kg/m Waist circumference, cm Systolic BP, mm Hg Diastolic BP, mm Hg NS Heart rate, bpm Baseline fasting glucose, mg/dl LVEF, % Total cholesterol, mg/dl Triglycerides, mg/dl LDL-C, mg/dl HDL-C, mg/dl Total/HDL-C BP indicates blood pressure; LVEF, left ventricular ejection fraction. diabetes was higher than in the nondiabetic group (149 versus 97 mg/dl). Pravastatin had similar effects on plasma lipid concentrations in the diabetes and nondiabetes groups (Figure 1). Compared with placebo, pravastatin treatment reduced total cholesterol and LDL-C by 19% and 27% in the diabetic patients and by 20% and 28%, respectively, in the nondiabetic patients. Average on-treatment total cholesterol and LDL-C values in the diabetes group were and mg/dl compared with and mg/dl in the nondiabetes group. Compared with placebo, pravastatin caused a 13% decrease in triglycerides and a 4% increase in HDL levels in the diabetic group, values which were similar to those obtained in the nondiabetic group (Figure 1). Cardiovascular events in the placebo group with and without a clinical diagnosis of diabetes are shown in Figure 2. Among the patients assigned to placebo, all the end points occurred with significantly greater frequency in the diabetes than in the nondiabetes group, except for angioplasty. The primary end point (CHD death or nonfatal MI) occurred in 20% of diabetic and 12% of nondiabetic patients (P 0.001). The expanded end point occurred in 37% of diabetic and 25% of nondiabetic patients (P 0.001). Stroke occurred in 8% of diabetic and 3% of nondiabetic patients assigned to placebo Figure 1. Effect of pravastatin on cholesterol and triglycerides in diabetes and nondiabetes groups. Change in pravastatin group is shown relative change in placebo group. Average for 5-year trial duration is used. P Figure 2. Cardiovascular event rates in patients with or without clinically diagnosed diabetes in placebo group (n 304 and 1774, respectively). Expanded end point was CHD death, nonfatal MI, CABG, or PTCA. *P 0.01; P (P 0.001). Adjustment of these incidence rates for differences in age or sex between the diabetes and nondiabetes groups did not affect the results, indicating that the slightly older mean age and higher proportion of women could not explain the higher event rate in the diabetes group. Pravastatin treatment in the diabetes group was associated with a 25% reduction of risk of coronary events (CHD death, nonfatal MI, CABG, and PTCA) (P 0.05), which was similar to that in the group without diabetes (Figures 3 and 4). Adjustment for age and sex did not alter the magnitude of the risk reduction. However, because of their higher event rate, the diabetes group experienced greater absolute risk reduction than the nondiabetes group (8.1% versus 5.2%). The diabetic patients in the pravastatin group had significantly fewer revascularization procedures (PTCA or CABG) than those in the placebo group (relative risk, 0.68; P 0.04). Pravastatin treatment was associated with a 26% reduction in the relative risk of primary end-point events in the nondiabetes group (P 0.004) and a 13% reduction in the diabetes group (P NS) (Figure 3). The small difference in the magnitude of the pravastatin effect on the primary end-point reduction in the 2 groups was due largely to the absence of a reduction in the rate of CHD death in the diabetes group. The risks for the individual cardiovascular events were consistently but not significantly lower in the pravastatin than the placebo group (Figure 3). There was no interaction between diabetes status and response to pravastatin in any of the end points. The associated risk for recurrent events according to cut points of baseline fasting glucose concentrations for the patients who did not have a clinical diagnosis of diabetes is shown in Figure 5. The relative risk of primary end-point events was assessed by comparing event rates in patients above versus below fasting glucose cut points beginning at 95 mg/dl and increasing at 5-mg/dL increments through 140 mg/dl (Table 3). A significantly greater relative risk for primary end-point events was noted among patients with

4 2516 Pravastatin, Diabetes, and CHD Figure 3. Effect of pravastatin on risks of cardiovascular events in patients with or without clinically diagnosed diabetes. Numbers indicate patients with cardiovascular events. Relative risks (RR) for pravastatin compared with placebo group are shown with 95% CIs. F F indicates total nondiabetes group; E- --E, diabetes group. fasting glucose values 110 mg/dl (relative risk, 1.41; P 0.01) compared with those at or below this glucose concentration. The results were similar for the expanded end point, or when the nondiabetic group was restricted to those whose baseline fasting glucose concentrations were 126 mg/dl (Figure 5). Adjustment for age and sex had no effect on the relative risks. Baseline characteristics in the nondiabetic group with baseline fasting glucose values of 110 to 125 mg/dl (n 342) were compared with those with values 110 mg/dl (n 3104). Those with glucose values of 110 to 125 mg/dl had larger waist circumferences and body mass indexes (BMIs), were older, were more likely to be hypertensive, and had higher baseline triglycerides and lower HDL concentrations than those with values 110 mg/dl (Table 4). There was no difference in their baseline LDL-C. There were no differences in the effects of pravastatin on plasma lipids or lipoprotein cholesterol fractions in the 2 groups (net average LDL-C reductions were 26% and 28% in the groups with glucose levels of 110 to 125 and 110 mg/dl, respectively). In these 342 patients without clinical diagnoses of diabetes with fasting blood glucose levels of 110 to 125 mg/dl, risks for most coronary events were lower in the pravastatin compared with the placebo group; eg, for the primary end point, the relative risk was 0.77; for clinically diagnosed nonfatal MI, the relative risk was 0.50; and for revascularization, the relative risk was But because of the small sample size, the differences were not statistically significant (Figure 6). These results were not affected by adjustment for age and sex. Discussion Although the association between diabetes and CHD has been clearly established, little information is available on the effects of interventions on cardiovascular events in these patients. By including patients with diabetes, the CARE trial afforded the opportunity to determine whether lowering LDL-C levels is beneficial in patients with diabetes, CHD, and usual levels of cholesterol. Furthermore, the availability of fasting glucose levels provided important information concerning the frequency and impact of recognized hyperglycemia in patients with established CHD and the benefits of pravastatin therapy in patients with impaired glucose tolerance. TABLE 3. Effect of Entry Glucose Cut Points on the Primary End Point in Nondiabetic Patients Glucose Cut Points, mg/dl n cut point n cut point Incidence rate cut point Incidence rate cut point Relative risk % 95% Confidence limits P Relative risk for primary end point in nondiabetic patients with fasting blood glucose above compared with below the indicated cut point.

5 Goldberg et al December 8, Figure 4. Kaplan-Meier plots for expanded coronary end point in placebo- and pravastatin-treated patients in diabetes and nondiabetes groups (n 586 and 3573, respectively). Expanded end point was CHD death, nonfatal MI, CABG, or PTCA. Fourteen percent of the total CARE population were known to be diabetic. Although patients with screening fasting glucose levels 220 mg/dl were excluded from the CARE trial by protocol, this prevalence falls within the range previously reported among patients with myocardial infarction. 15 History of diabetes in the CARE patients is likely to be reliable for the following reasons: 60% of the diabetic patients were treated with either an oral hypoglycemic drug or insulin, and only 14% of the diabetic group were not being treated and had a normal fasting glucose concentration ( 110 mg/dl). Many of these were likely to be diabetic patients whose glucose had been controlled by diet or weight management. Although some diabetic subjects have hypertriglyceridemia, the mean triglyceride concentration in a large population of diabetic subjects is only modestly higher than in nondiabetics. For example, in the National Health and Nutrition Examination Survey (NHANES II) survey, the mean triglyceride concentration in diabetics was 175 mg/dl 19 compared with 164 mg/dl in the diabetics in the CARE trial (in whom only those with the highest triglycerides, 350 mg/dl, were excluded). Thus, apart from a relatively small number of the most severely hyperglycemic and hypertriglyceridemic diabetic patients, the CARE diabetes population is representative of those identified in community-based studies. Absence of diabetes as determined by questioning the patients is likely to be accurate for several reasons. First, the patients had extensive medical evaluations as a result of their coronary disease that should have uncovered hyperglycemia. Only 3% of the nondiabetic group had a baseline fasting glucose measurement 125 mg/dl, suggesting the possibility of undiagnosed diabetes. We decided to leave this small number of patients in the nondiabetes group to ensure uniformity of the criteria for the subgroups by history of diabetes. We were also concerned that a single elevated fasting glucose concentration is not sufficient to diagnose diabetes at the same level of confidence as a clinical diagnosis, which is based on multiple measurements before and during treatment. Patients in the CARE trial with a history of diabetes were more obese and were more likely to have hypertension, cerebrovascular and peripheral vascular disease, and congestive heart failure than those without such a history, as has been reported previously. 20 Although there were relatively small numbers of patients from minority ethnic groups in CARE, they were disproportionately represented among individuals with diabetes, in keeping with the predisposition among these population groups to develop diabetes. 21 There was a greater proportion of women in the diabetic group than those without diagnosed diabetes, demonstrating again the importance of diabetes as a cardiovascular risk factor in women. 22 The diabetic group had a higher prevalence of hypertension, obesity, and congestive heart failure, which likely contributed to their higher event rate. Although the diabetic patients had average LDL-C levels, they had higher triglycerides and lower HDL-C, the typical dyslipidemia associated with diabetes. During the 5-year follow-up, diabetic patients experienced almost twice the cardiovascular disease end points as those without diabetes. This higher coronary event rate in diabetic patients was likely influenced by the presence of diabetes and the associated risk factor burden but was not a result of older age or more women in the diabetic group. The only exception was the 5-year incidence rate of PTCA, which was similar in the diabetes and nondiabetes groups and may reflect a lower incidence of 1-vessel coronary artery disease that was considered ideal for PTCA during the years that patients were followed in CARE (1989 to early 1996). Pravastatin treatment in the patients with diabetes was associated with a 25% relative reduction in risk for coronary events as measured by the expanded end point, similar to the Figure 5. Relative risk (RR) for recurrent coronary events according to cut points of baseline fasting glucose. Fasting glucose concentrations indicate cut points, each dividing population into those above and below indicated concentration. Relative risks denote risk of recurrent coronary event in patients above compared with below indicated fasting glucose concentration. Numbers of patients above and below each cut point are shown in Table 3 for total nondiabetes group. FBS indicates fasting blood sugar.

6 2518 Pravastatin, Diabetes, and CHD TABLE 4. Age, Glucose, BMI, Waist Circumference, Blood Pressure, and Lipid and Lipoprotein Levels at Baseline in Nondiabetic Patients With Fasting Glucose Tolerance Classification of Normal or Impaired Normal Baseline Glucose ( 110 mg/dl) Impaired Baseline Glucose ( mg/dl) P n Age, y Baseline glucose, mg/dl (By design) BMI, kg/m Waist circumference, cm Systolic BP, mm Hg Diastolic BP, mm Hg Total cholesterol, mg/dl Triglycerides, mg/dl LDL-C, mg/dl HDL-C, mg/dl Total/HDL-C % reduction found in the nondiabetes group. Because of their higher recurrent coronary event rate, diabetic patients had more benefit, expressed in absolute terms, than nondiabetic patients, an 8.1% absolute reduction of the expanded end point in the diabetic compared with 5.2% in the nondiabetic patients. The 4S demonstrated that simvastatin decreased major CHD events among a hypercholesterolemic diabetic group with diagnosed CHD. As in CARE, the percentage risk reductions were similar in diabetic and nondiabetic groups, but the absolute risk reduction was greater in the diabetic than nondiabetic patients. 10 The diabetic groups in 4S and CARE differed substantially in their baseline LDL-C (187 versus 136 mg/dl). In addition, HDL-C was higher in 4S than in CARE (44 versus 38 mg/dl), whereas triglycerides were lower (152 versus 164 mg/dl). Because of a more restrictive selection process, diabetic patients in 4S made up only 4.6% of the whole cohort compared with 14% in CARE. Compared with the patients in CARE, the diabetic patients in 4S were also less obese (BMI, 25.9 versus 29.4 kg/m 2 ), fewer were hypertensive (40% versus 52%) or taking aspirin (38% versus 78%), and more were taking -blockers (61% versus 39%). The results in CARE demonstrate that the typical diabetic patient with average LDL-C levels will benefit from further improvement in lipids with pravastatin treatment. The role of hyperglycemia as a risk factor for CHD among diabetic patients is controversial, 6 8 and there is no information on this subject in relation to recurrent CHD events. However, there is evidence that patients who are mildly glucose intolerant without evident CHD have an increased Figure 6. Effect of pravastatin on risks of cardiovascular events in nondiabetic patients with normal or impaired fasting glucose tolerance. Normal glucose tolerance was fasting blood glucose 110 mg/dl; impaired glucose tolerance, 110 to 125 mg/dl. Numbers indicate patients with cardiovascular events. Relative risks (RR) for pravastatin compared with placebo group is shown with 95% CIs.

7 Goldberg et al December 8, risk for coronary events. 9 This is important because most patients with impaired glucose tolerance and mild diabetes are undiagnosed. 23 The increased risk for CHD in patients with impaired glucose tolerance has been attributed to an increased preponderance of cardiovascular risk factors associated with this state. 24 Whatever the mechanism for the increased risk of CHD, previously unrecognized hyperglycemia constitutes a potentially important risk factor for recurrent cardiovascular events among patients with CHD. Although the diagnosis of impaired glucose tolerance and mild diabetes has traditionally required the performance of a cumbersome oral glucose tolerance test, we sought to determine whether there was a threshold fasting glucose level associated with an increased risk for CHD recurrences among CARE participants who had not been formally diagnosed with diabetes. With the use of the primary end point, successive testing at 5-mg/dL baseline fasting glucose increments revealed that the risk of CHD increased sharply and significantly at fasting blood sugars 110 mg/dl. It is of interest that the Expert Committee of the American Diabetes Association has recently redefined the criteria for the diagnosis of an abnormal fasting glucose value at 110 mg/dl in an effort to simplify the detection of undiagnosed glucose intolerance. 15 There were 342 patients (8% of all CARE patients) without a history of diabetes who had entry fasting glucose of 110 to 125 mg/dl, which corresponds to the American Diabetes Association definition of impaired fasting glucose. 15 These nondiabetic patients with impaired glucose tolerance resembled the diabetes group in having a more disadvantageous cardiovascular risk profile and were at increased risk for CHD events. Although this group was small and therefore there was insufficient power to assess adequately the effects of pravastatin treatment, the risk for most CHD end points was lower in the pravastatin compared with the placebo group. The percentage risk reductions in those with impaired glucose tolerance were similar to those observed in patients with fasting blood glucose 110 mg/dl. However, as in the diabetic patients, absolute risk reduction is greater in those with impaired compared with normal fasting glucose tolerance because of their high event rates. In summary, these findings demonstrate that pravastatin treatment significantly reduced the frequency of coronary events in diabetic patients with established CHD and average cholesterol levels. Thus, even though the predominant lipid problem in diabetic patients is high triglycerides and low HDL, a therapeutic strategy aimed at lowering LDL has major benefit. In addition, in nondiabetic patients, a fasting glucose level of 110 to 125 mg/dl identified a significant number of patients with undiagnosed impaired glucose tolerance who have an increased CHD risk that was ameliorated by pravastatin treatment and who therefore should be targeted for more aggressive risk factor management. References 1. Kannel WB, McGee DL. Diabetes and glucose tolerance as risk factors for cardiovascular disease: the Framingham Study. Diabetes Care. 1979;2: Wingard DL, Barrett-Connor E. Heart disease and diabetes. In: Harris MI, ed. Diabetes in America. 2nd ed. Bethesda, Md: National Institutes of Health, National Institutes of Diabetes and Digestive and Kidney Disease; 1995: NIH publication Leitersdorf E, Gottchrer N, Fainaru M, Friedlander Y, Friedman G, Tzivoni D, Stein Y. Analysis of risk factors in 532 survivors of first myocardial infarction hospitalized in Jerusalem. Atherosclerosis. 1976; 59: Pyörälä K, Laakso M, Uusitupa M. Diabetes and atherosclerosis: an epidemiologic view. Diabetes Metab Rev. 1987;3: Wilson PW, Kannel WB, Anderson KM. Lipids, glucose tolerance and vascular disease: the Framingham study. Monogr Atheroscler. 1985;13: Klein R. Hyperglycemia and microvascular and macrovascular disease in diabetes. Diabetes Care. 1995;18: West KM, Ahuja MMS, Bennett PH, Czyzyk A, De Acosta OM, Fuller JH, Grab B, Grabauskas V, Jarrett, RJ, Kosaka K, Keen H, Krolewski AS, Miki E, Schliack V, Teuscher A, Watkins, PJ, Stober JA. The role of circulating glucose and triglyceride concentrations and their interactions with other risk factors as determinants of arterial disease in nine diabetic population samples from the WHO multinational study. Diabetes Care. 1983;6: Ohlson L-O, Bjuro T., Larsson B, Eriksson H, Svardsudd K, Welin L, Wilhelmsen L. A cross-sectional analysis of glucose tolerance and cardiovascular disease in 67-year-old men. Diabetes Med. 1989;6: Lowe L. P., Liu K, Greenland O, Metzger BE, Dyer AR, Stamler J. Diabetes, asymptomatic hyperglycemia, and 22-year mortality in black and white men. Diabetes Care. 1997;20: PyöräläK, Pedersen TR, Kjeksus J, Faergeman O, Olsson AG, Thorgeirsson G. Cholesterol lowering with simvastatin improves prognosis of diabetic patients with coronary heart disease: a subgroup analysis of the Scandinavian Simvastatin Survival Study (4S). Diabetes Care. 1997;20: Sacks FM, Pfeffer MA, Moyé LA, Rouleau JL, Rutherford JD, Cole TG, Brown L, Warnica JW, Arnold JMO, Wun C, Davis BR, Braunwald E. The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels: Cholesterol and Recurrent Events Trial Investigators. N Engl J Med. 1996;335: Sacks FM, Pfeffer MA, Moyé L, Brown LE, Hamm P, Cole TG, Hawkins CM, Braunwald E. Rationale and design of a secondary prevention trial of lowering normal plasma cholesterol levels after acute myocardial infarction: the Cholesterol and Recurrent Events Trial (CARE). Am J Cardiol. 1991;68: Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Summary of the Second Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol (Adult Treatment Panel II). JAMA. 1993;209: Friedewald WT, Levy RI, Frederickson DS. Estimation of the concentration of low density lipoprotein cholesterol in plasma without use of the preparative ultracentrifuge. Clin Chem. 1972;18: Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care. 1997;20: Kalbfleisch ID, Prentice RI. The Statistical Analysis of Failure Time Data. New York, NY: John Wiley & Sons Inc; Cox DR. Regression models and life tables. J R Stat Soc (B). 1972;34: Sprafka IM, Burke GL, Folsom AR, McGovern, PG, Hahn LP. Trends in prevalence of diabetes mellitus in patients with myocardial infarction and effect of diabetes on survival: the Minnesota Heart Survey. Diabetes Care. 1991;14: Plan and operation of the Second National Health and Nutrition Examination Survey. Vital and Health Statistics. Washington, DC: National Center for Health Statistics; Series 1, No Cowie CC, Harris MI. Physical and metabolic characteristics of persons with diabetes. In: Harris MI, ed. Diabetes in America. 2nd ed. Bethesda, Md: National Institutes of Health, National Institutes of Diabetes and Digestive and Kidney Disease; 1995: NIH publication Harris MI. Epidemiological correlates of NIDDM in Hispanics, whites and blacks in the US population. Diabetes Care. 1991;14(suppl 3): Barrett-Connor EL, Cohn BA, Wingard DL, Edelstein SL. Why is diabetes mellitus a stronger risk factor for fatal ischemic heart disease in women than men? JAMA. 1991;265: Harris MI, Hadden WC, Knowler WC, Bennet PH. Prevalence of diabetes and impaired glucose tolerance and plasma glucose levels in US population aged years. Diabetes. 1987;36: Fuller JH, Shipley MJ, Rose G, Jarrett RJ, Keen H. Coronary heart disease and impaired glucose tolerance. Lancet. 1980;1: