Anthony B. Gustafson

Adiposity, Fat Distribution, and Cardiovascular Risk

STUDY OBJECTIVE To determine the relative importance of adiposity and fat distribution to cardiovascular risk profile. DESIGN A cross-sectional study. SETTING Clinical research center funded by the National Institutes of Health. PATIENTS Convenience sample of 33 healthy premenopausal women with a wide range of body weight who did not have diabetes mellitus, hirsutism and virilism, gynecologic disorder, cardiac disease, or hypertension. Women participating in exercise or dietary programs or taking medication were excluded. All subjects completed the study. INTERVENTIONS Total body fat mass was determined by hydrostatic weighting, and fat distribution was assessed by subscapular skinfold thickness, subscapular-to-triceps skinfold ratio, the waist-to-hip ratio, and computed tomography. Cardiovascular risk was assessed by the serum insulin response during oral glucose stimulation; levels of triglycerides and total cholesterol; high-density lipoprotein cholesterol to total cholesterol concentrations; and systolic and diastolic blood pressures. MEASUREMENTS AND MAIN RESULTS The anthropometric parameters chosen were significantly associated with the cardiovascular risk profile (P less than 0.001). Visceral fat distribution assessed by computed tomography accounted for a significantly greater degree of variance in the cardiovascular risk factors than the total body fat mass (P less than 0.05). The cumulative insulin response was the primary metabolic variable relating the anthropometric indices to cardiovascular risk. CONCLUSIONS Intra-abdominal fat deposition constitutes a greater cardiovascular risk than obesity alone. Hyperinsulinemia may constitute an important component of the increased cardiovascular risk of abdominal obesity.

Drug-Induced Disorders of Glucose Tolerance

Glucose concentrations are normally maintained between narrow limits through insulin secretion and action. The association of medication with alterations in glucose-insulin homeostasis is not new. Polypharmacy enhances the risk for drugdrug interactions and adverse drug effects. A study within our institution estimated the average number of medications per patient to be 4.5, with 12.9% of veterans taking more than eight prescriptions. Adverse drug effects may account for 3% to 5% of hospital admissions [1, 2]; approximately 50% of these admissions are preventable [2]. Between 30 000 and 140 000 patients have been estimated to have fatal drug reactions annually [3]. In hospitalized patients, adverse drug effects have been estimated to occur in about 30% of patients [4]. The economic costs of drug toxicity are an astonishing 3 to 4.5 billion dollars and are estimated to account for one seventh of all hospital days [5]. Drug-induced hypoglycemia may mimic disorders such as insulinoma and can lead to costly investigations. The concomitant use of drugs that may reduce blood glucose levels may induce profound hypoglycemia in patients receiving oral hypoglycemic agents or insulin. Furthermore, impaired glucose tolerance may enhance the risk for vascular disease [6]. Hyperinsulinemia and insulin resistance may be an intrinsic component of many disorders, such as hypertension, hyperlipidemia, and atherosclerosis. This combination of factors has been called Syndrome X by Reaven [7]. The increase in plasma insulin levels may promote these metabolic aberrations [8]. This review examines medications that may alter glucose insulin homeostasis and discusses possible mechanisms of action. Figure 1 shows the potential sites at which medication may induce changes in glucose metabolism. Figure 1. Potential sites of action for drugs influencing glucose metabolism. Medications Associated with Hyperglycemia Thiazide Diuretics The adverse metabolic effects induced by thiazide diuretics have been implicated in the failure to dramatically reduce coronary artery disease risk, despite a significant reduction in cerebrovascular disease and congestive cardiac failure [9]. Thiazide diuretics have been implicated as factors in inducing glucose intolerance in nondiabetic and diabetic patients in many studies. A 30% incidence of glucose intolerance in hypertensive patients receiving thiazide diuretics has been reported [10]; however, it is important to realize that glucose intolerance from thiazides may not be immediately apparent [11]. Higher doses are more likely to be associated with glucose intolerance. A direct toxic effect on the pancreas has also been postulated [12]. Diuretic-induced hyperglycemia may be due to decreased insulin secretion as a result of hypokalemia [13]. The reduction in total body potassium correlates with a reduction in insulin secretion. Furthermore, correction of hypokalemia by replacement with potassium salts can prevent the deterioration in glucose tolerance and may restore insulin sensitivity [14]. Another possible contributor to elevated glucose levels may be enhanced free fatty acid and lipid exposure of tissues subsequent to thiazide use [9]. Other mechanisms that may result in hyperglycemia include decreased insulin sensitivity, increased hepatic glucose production, a direct inhibitory effect on insulin secretion, enhanced catecholamine secretion and action, and phosphodiesterase inhibition [15-17]. Apart from possible effects on cells, a stimulatory effect on cells has also been described in association with thiazides [18]. It is not known whether glucagon secretion is enhanced to a clinically relevant degree. Metabolic adverse effects may vary between drugs of this group. Chlorthalidone may be more likely than hydrochlorothiazide to induce hypokalemia [19]. Indapamide was reported to be less likely to induce abnormalities in glucose metabolism [20]; however, more recent studies on indapamide have not confirmed the lack of deleterious effects on glucose tolerance [21]. Glucose intolerance occurs less commonly with loop diuretics [22]. Hypokalemia may contribute to a diabetogenic tendency in some instances. In addition, furosemide may cause glucose intolerance in some patients by decreasing insulin release, possibly through increased synthesis of prostaglandin E [23]. Of the loop diuretics, ethacrynic acid is least likely to have a diabetogenic effect. Potassium-sparing diuretics such as spironolactone or triamterene have minimal or no effects on glucose tolerance [24]. Centrally Acting -Blockers Early animal studies suggested a possible hyperglycemic effect of these agents. Clonidine may reduce insulin secretion and thus impair glucose tolerance in diabetic patients [25]. Alpha receptor suppression of insulin release as well as possible central effects may provide a theoretic basis for anticipating altered glycemia. In practice, however, no significant glucose intolerance has been found in most studies in humans [26]. Beta-Blockers Beta-blockers may inhibit insulin release [27], and hyperosmolar coma has been associated with the use of these agents. A greater inhibitory effect on insulin secretion may be observed with nonselective -blockers [28]. Lipophilicity may have a greater adverse effect than nonselectivity on plasma glucose values [29, 30]. Reduced insulin secretion may lead to enhanced hepatic glucose production, thereby contributing to glucose intolerance. A decrease in hepatic and peripheral glucose uptake may occur after use of these agents [31]. Blockage of other -receptor-mediated effects such as glycogenolysis in muscle may also influence plasma glucose levels. In general, marked hyperglycemia is uncommon with the use of these agents. It is important to note that -blockers may also cause hypoglycemia (see discussion of hypoglycemia). Calcium-Channel Blockers Insulin release depends on an increase in cytosolic calcium in vitro [32] and calcium-channel blockers have been used in the treatment of insulinoma. Early reports suggested that calcium-channel blockers may reduce insulin secretion and induce hyperglycemia in humans [33, 34]. The deleterious effects of calcium-channel blockers on insulin secretion may be dose dependent [35]; however, adverse effects of nifedipine on carbohydrate metabolism have not been found by other investigators [36]. In some reports, verapamil has improved glucose tolerance by decreasing glucagon release and by enhancing hepatic glucose uptake [37]. Others have found a marked hyperglycemic effect with therapeutic doses of verapamil [38]. Diltiazem may have less marked effects on carbohydrate metabolism [39], although an increase in insulin requirements in a patient with type 1 diabetes has been reported [40]. Although the potential to induce glucose intolerance does exist with these agents, it appears that clinical use is generally not accompanied by severe hyperglycemia. Minoxidil Plasma glucose values can increase with the use of minoxidil in diabetic patients [41]. The use of minoxidil in 13 patients with advanced hypertension resulted in the development of diabetes in 1 patient. In two other patients with diet-controlled diabetes, oral hypoglycemic medication had to be added to their regimens. Diazoxide The use of this agent in the short-term treatment of severe hypertension has been associated with hyperglycemia and hyperosmolar nonketotic coma. Its hyperglycemic effects are observed more frequently and are more severe than those seen with thiazide agents. Direct inhibition of insulin secretion may be the primary mechanism responsible for the hyperglycemia [42]. Patients with insulinoma have previously been treated with this agent [16]. Other mechanisms implicated include direct stimulation of hepatic glucose production, increased epinephrine secretion, decreased insulin sensitivity, and increased insulin clearance [43]. Corticosteroids Long and colleagues [44] showed the diabetogenic effect of glucocorticoids in 1940 and attributed this effect to enhanced gluconeogenesis. More recent evidence indicates that therapeutic doses of glucocorticoids may also impair glucose uptake [45]. Insulin resistance appears to occur at both receptor and postreceptor sites [46, 47], and variations between glucocorticoids with regard to insulin binding do exist [48]. Healthy persons may regain normal glucose tolerance even if the drug is continued. In patients with diabetes or impaired glucose tolerance, however, the diabetogenic effects can be prolonged, and hyperosmolar nonketotic coma has been reported [49]. A rapid increase in plasma glucose (within 24 hours) can be seen with these agents. The effects of glucocorticoids on carbohydrate metabolism in susceptible patients are dose related and are most often seen with the systemic use of these drugs [16]. Topical glucocorticoid use, however, can be associated with glucose intolerance [50, 51]. This result is more likely if more potent steroids are used for prolonged periods over a large surface area and with the use of occlusive dressings. Recovery is usually prompt after discontinuation of the drug. Although all glucocorticoids can induce glucose intolerance, the glucocorticoids that are oxygenated in the 11- and 17-positions, such as hydrocortisone and the presence of a 1,2 double-bond in the A ring (prednisone and prednisolone), have the most diabetogenic effects. Glucocorticoids can also induce hyperglycemia through stimulation of the cells, leading to hyperglucagonemia and increased glycogenolysis [43]; other mechanisms include increased gluconeogenesis. Alternate-day steroid use to ameliorate the hyperglycemic effect of steroids has resulted in alternate-day hyperglycemia [52]. Adrenocorticotropic hormone has a diabetogenic effect similar to that of glucocorticoids. Mineralocorticoids do not directly influence carbohydrate metabolism, although hypokalemia associated with the use of these agents may reduce insulin secretion

Enkephalins, catecholamines, and psychological mood alterations: effects of prolonged exercise

Seven healthy trained men were studied to determine if running at various relative intensities [percent maximal oxygen consumption (VO2max)] alters peripheral venous levels of leucine enkephalin-like material. Enkephalins were measured using a radio-receptor assay (Leu-Enk RRA). Subjects ran for 80 min at 40 and 60% VO2max and for 40 min at 80% VO2max. Each session was separated by at least 1 wk. Heart rate, blood pressure, lactic acid, and rectal temperature responses increased in an intensity-dependent manner. Epinephrine increased from resting values of 38.2 +/- 6.8 pg X ml-1, mean +/- SE to 75.0 +/- 13.3 pg X ml-1 during the 40% VO2max run, from 60.2 +/- 15 to 186 +/- 45 pg X ml-1 during the 60% run, and from 33.4 +/- 7.6 to 311 +/- 52 pg X ml-1 at the 40th min of the highest workload (80% VO2max). These increases were significant (P less than 0.05). Plasma Leu-Enk RRA was between 3.8 and 6.2 pmol X ml-1 prior to each run and did not change significantly as a result of exercise. Levels of Leu-Enk RRA also did not change during 30 min of supine recovery. Perception of effort increased (P less than 0.05) with increases in exercise intensity, and effort sense was unrelated to plasma Leu-Enk RRA. Psychological tension decreased significantly (P less than 0.05) following exercise at 60 and 80% of VO2max, but the decrease following the 40% run was not significant (P greater than 0.05). Reduced tension following exercise was not related to Leu-Enk RRA.(ABSTRACT TRUNCATED AT 250 WORDS)

Exercise Stress and Endogenous Opiates

The use of exercise to reveal physiological responses and adaptations has a long history. Exercise can be quantified and repeated exertion markedly alters body functions. Well established exercise procedures such as the use of an individual’s aerobic capacity have allowed exercise stress to be used as an experimental model to investigate the endogenous opiate system. A vast literature had accumulated prior to 1975 concerning the physiological effects of morphine. When the endogenous opiates were discovered’, researchers already had insights into probable functions of the endogenous opiates. Therefore, it is not surprising that many possible roles for endorphins/enkephalins have been studied using the exercise model.