Interrelationships of Hyperinsulinism and Hypertriglyceridemia in Young Patients with Coronary Heart Disease

Transcription

1 Interrelationships of Hyperinsulinism and Hypertriglyceridemia in Young Patients with Coronary Heart Disease By MANUEL TZAGOURNIS, M.D., Ross CHILES, M.D., JOSEPH M. RYAN, M.D., AND THOMAS G. SKILLMAN, M.D. SUMMARY Fasting serum lipid levels, glucose tolerance, and immunoreactive insulin concentrations of 50 young patients with coronary heart disease (CHD) and 30 control subjects were evaluated to study the interrelationships of these metabolic factors. Abnormalities in one or more of these factors could be shown in 90% of the patients and 20% of the control subjects. Thirty-four of the 50 patients had elevated cholesterol or triglyceride levels, or both, 30 had abnormally elevated or delayed insulin responses after glucose, and 17 had abnormal glucose tolerance. A significant correlation existed between serum triglyceride and insulin concentrations. When insulin levels were reduced by phenformin, triglyceride concentrations fell toward normal. These findings indicate that carbohydrate, insulin, and lipid abnormalities are rather prevalent in patients with CHD. Excessive insulin secretion secondary to mild glucose intolerance probably induces hepatic synthesis of triglycerides and hypertriglyceridemia. Dietary alterations or pharmacological agents may help to control some of the metabolic abnormalities associated with premature CHD. Additional Indexing Words: Atherosclerosis Diabetes Hyperglycemia Hyperlipidemia RECENT STUDIES suggest that atherosclerosis may involve a continuum in which hyperlipidemia and carbohydrate intolerance exercise pathophysiological significance. Patients with coronary heart disease (CHD) are commonly found to have elevated serum cholesterol and triglyceride concentrations.' 2 Many subjects with CHD also have abnormal glucose tolerance and immunoreactive insulin responses after carbohydrate loading.3 4 Although subjects with clinical diabetes mellitus as a group develop premature anti From the Divisions of Endocrinology and Metabolism, and Cardiology, Department of Medicine, The Ohio State University College of Medicine, Columbus, Ohio. Supported in part by U. S. Public Health Service Training Grant T01-AM05118 and U. S. Public Health Service General Research Grant 5-S01-FR , and the U. S. Vitamin and Pharmaceutical Corp extensive atherosclerosis, hyperglycemia, the traditional identifying metabolic marker for the diabetic state, does not seem to be directly related to the pathogenesis of atherosclerosis. Hyperlipidemia, on the other hand, may well play an important role in the genesis of the atherosclerotic lesion. The hormone insulin is fundamental to the attainment of both glucose and lipid homeostasis. The serum immunoreactive insulin response to orally administered glucose in maturity-onset diabetes is characterized by a delayed secretion, and the serum insulin levels are often elevated. Patients with hypertriglyceridemia also have been reported to have impaired glucose tolerance and elevated plasma insulin responses.5, 6 Farquhar and associates7 have reported that elevations of plasma glucose and insulin concentrations correlate well with the magnitude of the plasma triglyceride increase after

2 HYPERINSULINISM AND HYPERTRIGLYCERIDEMIA 1157 Table 1 Characteristics and Laboratory Data of CHD and Control Patients Age Patients (yr) Sex TG* Cholf IA* Coronary heart disease patients 30 M M M M M F M M M M M M M M M M M M M M M M M M M M M M M M M M M M M F M M M M M M M F M M M M F M Circulation, Volume XXXVIII. December 1968 Table 1 (continued) Age Patients (yr) Sex TG* Cholt Control patients 1 30 M M M M M F M M M M M M F M F M M M M M F M M M M M M M M m *Triglyceride level in mg%. IA* tcholesterol level in mg%. *Area under 3-hour insulin curve in milli-unit minutes. carbohydrate feeding. They suggested that elevated insulin levels in the presence of normal or high glucose levels enhance hepatic production of triglycerides. This view is consistent with other clinical studies indicating that the intake of sucrose and other simple carbohydrates, in contrast to complex carbohydrates such as starch, induces a sharp increase in blood glucose, insulin, and triglyceride concentrations.8-10 Eliminating the simple sugars usually brings the serum triglycerides back to normal. In this report various measurements were made to study the interrelationships of carbohydrate, insulin, and lipid metabolism in

3 1158 TZAGOURNIS ET AL. young patients with CHD. As a group, these patients displayed decreased glucose tolerance, abnormal insulin secretion, and elevated serum cholesterol and triglyceride concentrations. In addition, serum insulin levels correlated well with the fasting serum triglyceride concentrations. Methods Fifty patients with coronary heart disease (CHD) between the ages of 26 and 49 years were compared with 30 control subjects (table 1). Young patients were chosen because they might be expected to show certain metabolic abnormalities related to the development of atherosclerosis. The mean age of the group with CHD was 40 years and that of the control group, 36 years. The criteria for the diagnosis of coronary heart disease were: (1) definite myocardial infarction, confirmed by electrocardiographic changes or serum enzyme elevations or both, (36 patients); (2) angina pectoris with angiographic evidence of coronary atherosclerosis (seven patients); and (3) angina pectoris associated with electrocardiographic evidence of coronary insufficiency (seven patients). Known diabetics were excluded from the study. Since obesity is associated with increased serum insulin levels, control subjects as well as patients exceeding 115% of ideal body weight (as indicated in the Metropolitan Life Insurance Company weight tables) were also excluded. The patients with myocardial infarctions were not studied until at least 1 month following recovery from the acute episode. Patients and control subjects were ambulatory at the time of the investigation and were instructed to eat a normal diet with ample carbohydrate for 2 days preceding the testing. Serum cholesterol was determined according to the method of Abell and coworkers'1; serum phospholipids, by the method of Lowry and associates.12 Serum triglycerides were calculated from serum esterified fatty acids by the technique of Stern and Shapiro.'3 A 3-hour glucose tolerance test was performed with sampling at 0,,lk, 1, 2, and 3 hours for both glucose and insulin after 75 g of glucose had been administered. Determinations were made for glucose on whole blood by the Hoffinan method on the autoanalyzer'4 and serum immunoreactive insulin by the technique of Grodsky and Forsham.'5 Table 2 The Various Abnormalities Found in the Tested Subjects To obtain further information concerning lipid and insulin interrelationships, 30 patients were given phenformin, an oral hypoglycemic agent, which has also been shown to reduce serum insulin levels in obese and diabetic patients.16 The dosage ranged from 50 mg to 150 mg per day. After treatment for approximately 2 months, the serum lipids, glucose tolerance, and serum insulin measurements were repeated. Serum lipid determinations were subsequently performed at approximately 3-month intervals for 1 year. Results Frequency of Abnormal Findings in Patients with Coronary Heart Disease and Control Subjects Abnormalities of the serum lipids, serum insulin response, or glucose tolerance were 50 Patients 30 Control Abnormalities* with CHD subjects Insulin, cholesterol, and triglyceride 8 Insulin and cholesterol 5 Insulin only 5 1 Cholesterol only 5 2 Cholesterol and triglyceride 4 Insulin, carbohydrate, and triglyceride 4 1 Insulin, carbohydrate, and cholesterol 3 Insulin and carbohydrate 3 Carbohydrate only 3 Insulin, carbohydrate, cholesterol, and triglyceride 2 Triglyceride only 1 2 Triglyceride and carbohydrate 1 Cholesterol and carbohydrate 1 Total patients with abnormalities 45 (90%) 6 (20%) *Abnormal values were: Cholesterol-serum levels exceeding 290 mg%; triglycerides-serum levels exceeding 170 mg%; insulin-response manifested by elevated levels or by delayed secretion; carbohydrate-abnormal glucose tolerance test.

4 HYPERINSULINISM AND HYPERTRIGLYCERIDEMIA 1159 significantly greater in the patients with CHD than in the control subjects. One or more metabolic abnormalities could be demonstrated in 45 (90%) of the 50 patients and in six (20%) of the 30 control subjects (table 2) Ȧn elevation of the serum lipid levels was found in 34 of the 50 patients. The lipid levels were considered to be elevated if the serum cholesterol was greater than 290 mg% or the serum triglycerides was greater than 170 mg%. Fourteen patients had serum cholesterol concentrations exceeding 290 mg% and normal triglyceride levels. Six patients showed an elevated serum triglyceride value with a normal cholesterol concentration. The remaining 14 patients had an elevation in both serum cholesterol and triglyceride levels. Twentyfour of the 34 patients with elevated serum lipids also had an abnormal serum insulin response or impaired glucose tolerance, or both. The mean cholesterol value of the 50 patients was mg% (one standard deviation) as compared to 222±46 mg% in the control subjects. The mean triglyceride concentration was mg% in the patient group and mg% in the control group (table 3). An abnormal serum insulin response, manifested by elevated levels or delayed peak levels occurring at the 2 or 3-hour interval, was found in 30 of the 50 patients. The serum insulin level was considered to be elevated if it exceeded the mean plus two standard deviations of the control group's value for any given interval during the glucose tolerance test. The mean insulin response of the patients after glucose was higher at each interval than that observed in the control subjects (table 3). The total mean insulin concentration of the 50 patients, expressed as the area encompassed by the 3-hour curve, was mu-min (milli-unit minutes). This was significantly higher than the mu-min found in the 30 control subjects (P <0.05). Although no patient had known diabetes mellitus prior to the studies, 17 of the 50 patients had an abnormal glucose tolerance test according to the criteria of Wilkerson.17 The relative mildness of the abnormal glucose tolerance is reflected by the fact that no fasting blood glucose level exceeded 100 mg%. The mean glucose values at the and 2-hour intervals were, however, higher in the patient group than in the control subjects (table 3). Correlation of Serum Triglyceride Measurements with Other Variables in the Patients with Coronary Heart Disease (Table 4) A correlation between the fasting serum triglyceride concentration was sought with the following: (1) peak insulin level; (2) total insulin area under the 3-hour curve (mu-min); (3) serum cholesterol concentration; (4) 2-hour glucose level and (5) the percentage of ideal body weight. Fasting triglyceride values correlated well with the area under the 3-hour insulin curve (r = 0.34, P < 0.02). Figure demonstrates the tendency for the insulin concentrations to increase as the serum triglyceride levels rose. Correlations of serum triglyceride levels were also good with the peak insulin level (r = 0.28, P < 0.05) and the serum cholesterol Table 3 Comparison of Mean Lipid Glucose and Insulin Concentrations in 50 Patients with Coronary Disease and 30 Control Subjects Choles- Triglycterol eride Glucose (mg%) Insulin (,uu i ml) (mg%) (mg%) F '2 hr 1 hr 2 hr 3 hr F /2 hr 1 hr 2 hr 3 hr Controls Mean SD Coronary patients Mean SD

5 ISTANDARD ERROR TZAGOURNIS ET AL. Meon Voluev of 30 patiels befor* treatment After 2mosthi treatment with pheormls I STANDARD ERROR 20 I- Ci' E 100- w z N 13 N 13 N=12 N=12 First Quartile Second Quartile Third Quartile FourthQuartile (O -100) ( ) ( ) (>235) TRIGLYCERIDES mg.% Figure 1 Mean fasting serum concentrations of triglycerides in each quartile plotted against the mean insulin secretion (mu-min) of 50 patients with coronary heart disease. level (r = 0.31, P <0.05). No significant correlation existed between triglycerides and either the 2-hour glucose concentrations (r= 0.11, P>0.10) or percentage of ideal body weight (r=0.19, P>0.10). The poor relationship between the triglyceride level and body weight may have been due to the exclusion of individuals exceeding 115% of ideal body weight. Effects of Treatment with Phenformin Thirty of the 50 patients, all of whom had one or more abnormalities of their carbohydrate, insulin, or lipid metabolism, agreed to ll. a z HOURS Figure 2 The mean and one standard error of glucose and insulin values in 30 patients with CHD compared to values obtained after phenformin treatment for approximately 2 months. take phenformin and be re-evaluated at regular intervals. After treatment with phenformin, 50 to 150 mg per day for approximately 2 months, the serum insulin levels decreased in 26 of 30 patients. Figure 2 demonstrates that the mean insulin response was lower at each corresponding interval after glucose than that observed in the same Table 4 Summary of the Correlations (r) of Mean Values Among the 50 Patients with CHD Insulin Peak 2-Hour % Ideal Triglyceride area insulin Cholesterol glucose body weight Triglyceride 0.34* 0.28* 0.31* Insulin area 0.34* 0.90* Peak insulin 0.28* 0.90* Cholesterol 0.31* Hour glucose % Ideal body weight *P <0.05.

6 HYPERINSULINISM AND HYPERTRIGLYCERIDEMIA 25 1 E 20- E z 15-0 hi w 10 - z z BEFORE AFTER Phenformin 2 months INSULIN I STANDARD ERROR BEFORE AFTER AFTER Phenformin 2montt year Figure 3 TRI GLYCERI DES X 150 OCa -125 i -100 CD tr -75 *50 t -25 ) Response of insulin secretion (reflected by the area under the 3-hour curve) and fasting serum triglyceride levels to phenformin in 30 patients with coronary heart disease. patients prior to treatment. The mean of the areas under the 3-hour insulin curves decreased significantly from mu-min to mu-min (P < 0.05).. As expected, the levels of the glucose obtained during the tolerance tests decreased after treatment (fig. 2). The decrease in insulin secretion was accompanied by a fall in the level of serum triglycerides (fig. 3). The mean pretreatment serum triglyceride concentration of 191 mg% decreased to 142 mg% after 2 months. Sequential follow-up serum lipid concentrations at 3-month intervals were obtained on 25 of the 30 patients and the mean triglyceride level decreased to 127 mg% after 12 months of treatment (P <0.05). There was no significant change in body weight during the treatment. The mean initial weight of 161 lb decreased to 160 lb after 2 months and to 155 lb after 12 months of treatment. No attempt was made to alter the diet which the patients were taking at the time that they were referred for the study program. Discussion Despite the identification of numerous metabolic abnormalities associated with CHD, it is not clear whether any one factor is directly or independently related to the development of coronary atherosclerosis. For CirVdation, Volume XXXVIII, pecember 1, example, it is not known whether hyperglycemia alone plays a major role in atherogenesis, or whether its presence promotes abnormal lipid metabolism which then is basic to the atherosclerotic process. Therefore, it is useful to study the interrelationships of the metabolic abnormalities found in patients with CHD. This study of 50 patients with premature coronary atherosclerosis disclosed distinct abnormalities of glucose, lipid, and insulin metabolism in 90% of the group. Since the subjects were selected because they had definite CHD, the finding of elevated serum triglyceride and cholesterol concentrations was not surprising." 18 The high prevalence of abnormal glucose tolerance tests was somewhat unexpected because subjects with known clinical diabetes and elevated fasting blood glucose levels were deliberately excluded. This study also disclosed a significant positive correlation between fasting serum triglyceride concentrations and the magnitude of glucose-induced insulin secretion as well as the level of serum cholesterol. The association of hypertriglyceridemia and hypercholesterolemia, commonly found in our patients, is not unusual. Endogenously synthesized triglycerides are transported in the blood chiefly as pre-beta lipoproteins. These lipoproteins also transport cholesterol and, when increased, are associated with relative hypercholesterolemia. Although the association of a constellation of abnormal metabolic characteristics does not necessarily imply a causal relationship between them, there is good evidence to support the concept that insulin hypersecretion may provoke hypertriglyceridemia. When healthy individuals are fed a diet very high in carbohydrate, hypertriglyceridemia can be induced in them.'9 Furthermore, when subjects with elevated triglyceride levels are fed a diet which is poor in simple sugars, their serum lipid concentration almost always falls.8 It has also been shown that healthy persons secrete more insulin in response to feeding of the simple sugars, glucose and sucrose, than to a calorically equal amoiunt

7 1162 of starch.'0 These observations support the hypothesis of Reaven and associates20 which states that increased hepatic synthesis of triglycerides and release may play a causal role in the establishment of hypertriglyceridemia. Insulin plays an important role in the "fed state" in storing calories not only as glycogen, but also as protein and fat. It appears probable that impairment of glucose uptake by muscle and adipose tissue permitted an augmented insulin secretion in the CHD patients. Although the patients with mild glucose intolerance released less insulin for the degree of hyperglycemia present,2' their absolute output of insulin was greater than in the normal subjects. The observed excess of insulin, although insufficient to permit normal glucose uptake by cells, could have been adequate to permit an increase in hepatic triglyceride synthesis with subsequent hypertriglyceridemia. The conversion of glucose to triglycerides by the liver is, of course, not the only source of this lipid in man. Primary fat-induced lipemia, in contrast to "ccarbohydrate-induced" lipemia, has its origin in exogenous or dietary fat. In exogenous hypertriglyceridemia there is an inherited or acquired inability to dispose of chylomicrons, presumably due to a deficiency of the enzyme, lipoprotein lipase.22, 23 Still another source of triglyceride is hepatic conversion of free fatty acids (FFA) liberated by adipose tissue. A variety of stimuli may cause FFA mobilization in excess of the liver's capacity to oxidize them and therefore, facilitate their conversion to triglycerides. This mechanism may be operating in the relatively severe diabetic who has insufficient insulin for inhibition of adipose tissue lipolysis. This defect permits exaggerated FFA mobilization and increased hepatic triglyceride conversion. It seems, then, that insulin plays a major role in regulating normal lipid metabolism and that either abnormally increased or deficient amounts of this hormone can result in hypertriglyceridemia. Patients with CHD have been shown to have normal fasting levels of FFA TZAGOURNIS ET AL. as well as an adequate decline in FFA concentrations after a glucose load.4 It is, therefore, likely that both peripheral lipolysis and lipogenesis are normal in CHD patients, and that the hypertriglyceridemia is related to insulin-dependent, carbohydrate-induced hepatic synthesis. We hypothesized that a reduction of insulin secretion might be attended by an associated reduction in triglycerides. The biguanide, phenformin, has been demonstrated to decrease serum insulin levels in both obese and diabetic subjects.'6 Accordingly this agent was given to 30 of the CHD patients and follow-up studies were instituted. The evidence indicated that phenformin-induced decreases in blood glucose and insulin levels were associated with a fall in serum triglyceride concentrations toward normal. It is noteworthy that the lower levels persisted for a relatively long period with continued treatment. The close interrelationships of carbohydrate, insulin, and lipid metabolism make it difficult to estimate the importance of any single abnormality in the process of atherogenesis. Because atherosclerotic lesions in man contain triglycerides, cholesterol, phospholipids, and other substances such as mucopolysaccharides and fibrous tissue, emphasis has been placed on serum lipids in the pathogenesis of the disease. Certainly, individuals with elevated cholesterol levels without associated carbohydrate, triglyceride, or insulin abnormalities develop premature and extensive CHD. On the other hand, it is important to identify as many biochemical abnormalities as possible in a disease of unknown etiology. It is possible, for example, that exposure of arterial intima to elevated levels of insulin over a prolonged period exerts changes conducive to lipid accumulation by this tissue. Mahler24 has demonstrated that insulin not only stimulates lipid synthesis in arterial tissue by in vitro techniques, but also inhibits lipolytic activity. Both of these actions of insulin would favor lipid accumulation in the intima. Circulation, Volu7ne XXXVIII, December 1968

8 HYPERINSULINISM AND HYPERTRIGLYCERIDEMMIA 1163 An understanding, therefore, of the interrelationships of insulin and triglyceride metabolism with other biochemical changes may be useful in studying the pathogenesis of the disease process as well as in identifying susceptible individuals. This concept is consistent with the hypothesis that dietary alteration, such as decreased simple carbohydrate intake, or pharmacological agents may control some of the metabolic abnormalities associated with atherosclerosis. References 1. KANNEL, W. B., DAWBER, T. R., FRIEDMAN, G. D., GLENNON, W. E., AND MCNAMARA, P. M.: Risk factors in coronary heart disease: Evaluation of several lipids as predictors of coronary heart disease, Framingham study Ann Intern Med 61: 888, OSTRANDER, L. D., NEFF, B. J., BLOCK, W. D., FRANCIS, T., JR., AND EPSTEIN, F. H.: HYperglycemia and hypertriglyceridemia amona persons with coronary heart disease. Ann Intern Med 67: 34, PETERS, N., AND HALES, C. N.: Plasma insulin concentrations after myocardial infarctions. Lancet 1: 1144, TZAGOURNIS, M., SEIDENSTICKER, J. F., AND HAMWI, G. J.: Serum insulin, carbohydrate, and lipid abnormalities in patients with premature coronary heart disease. Ann Intern Med 67: 42, ALBRINK, M. J., AND DAVIDSON, P. C.: Impaired glucose tolerance in patients with hypertriglyceridemia. J Lab Clin Med 67: 573, DAVIDSON, P. C., AND ALBRINK, M. J.: Abnormal plasma insulin response with high plasma triglycerides independent of clinical diabetes or obesity. J Clin Invest 45: 1000, FARQUHAR, J. W., FRANK, A., GRoss, R. C., AND REAVEN, G. M.: Glucose, insulin, and triglyceride responses to high and low carbohydrate diets in man. J Clin Invest 45: 1648, Kuo, P. T.: Hyperglyceridemia in coronary artery disease and its management. JAMA 201: 87, COHEN, A. M., KAUFMANN, N. A., POZNANSKI, R., BLONDHEIM, S. H., AND STEIN, Y.: Effect of starch and sucrose on carbohydrate-induced byperlipemia. Brit Med J 1: 339, SWAN, D. C., DAVmSON, P., AND ALBRINK, M. D.: Effect of simple and complex carbohydrate on plasma nonesterified fatty acids, plasma sugar, and plasma insulin during oral carbohydrate tolerance tests. Lancet 1: 60, ABELL, L. L., LEVY, B. B., BRODIE, B. B., AND KENDALL, F. E.: Simplified method for estimation of total cholesterol in serum and demonstration of its specificity. J Biol Chem 195: 357, LOWRY, 0. H., ROBERTS, N. R., LEINER, K. T., Wu, M. L., AND FARR, A. L.: Quantitative histochemistry of brain: Chemical methods. J Biol Chem 207: 1, STERN, I., AND SHAPIRO, B.: Rapid and simple method for determination of esterified fatty acids and for total fatty acids in blood. Brit J Clin Path 6: 158, HOFFMAN, W. S.: Rapid photoelectric method for determination of glucose in blood and urine. J Biol Chem 120: 51, GRODSKY, G. M., AND FORSHAM, P. H.: Immunochemical assay of total extractable insulin in man. J Clin Invest 39: 1070, GRODSKY, G. M., KARAM, J. C., PAVLATOS, F. C., AND FORSHAM, P. H.: Reduction by phenformin of excessive insulin levels after glucose loading in obese and diabetic subjects. Metabolism 12: 278, WILKERSON, H. L. C.: Diagnosis: Oral glucose tolerance tests. In Diabetes Mellitus: Diagnosis and Treatment, edited by T. S. Danowaski. Amer Diabetes Assoc., Inc., N. Y., 1964, pp ALBRINK, M. J., MEIGS, J. W., AND MAN, E. B.: Serum lipids, hypertension, and coronary artery disease. Amer J Med 31: 4, Kuo, P. T.: Dietary sugar in the production of hypertriglyceridemia in patients with hyperlipemia and atherosclerosis. Trans Ass Amer Physicians 78: 97, REAVEN, C. M., LERNER, R. L., STERN, M. P., AND FARQUHAR, J. W.: Role of insulin in endogenous hypertriglyceridemia. J Clin Invest 46: 1756, SELTZER, H. S., ALLEN, E. W., HERRON, A. L., JR., AND BRENNAN, M. T.: Insulin secretion in response to glycemic stimulus: Relation of delayed initial release to carbohydrate intolerance in mild diabetes mellitus. J Clin Invest 46: 323, HAVEL, R. J., AND GORDON, R. S., JR.: Idiopathic hyperlipidemia: metabolic studies in an affected family. J Clin Invest 39: 1777, BAGDADE, J. D., PORTE, D., JR., AND BIERMAN, E. L.: Diabetic lipemia: Form of acquired fat-induced lipemia. New Eng J Med 276: 427, MAHLER, R.: Diabetes and arterial lipids (abst.). Quart J Med 34; 484, 1965.