Fig. 2. Leg blood flow under basal conditions (saline), in response to 4 hours of euglycemic hyperinsulinemia alone (insulin) and with superimposed intrafemoral artery infusion ofl-NMMA (insulin + l-NMMA). (From ref. 11a.)

(Fig. 2) levels within 5 minutes of an infusion of l-NMMA into the femoral artery (11). Our findings have been confirmed by others in humans (15) and in animals (16). Thus, it is now well established that insulin increases skeletal muscle blood flow via release of endothelial-derived NO.

The notion that insulin acts via release of NO from endothelial cell is supported by the observation that insulin directly releases NO from human umbilical vein endothelial cells (17). This insulin-mediated NO release occurred in a dose-dependent fashion and could be completely abolished by ^(omega)-nitro-L-arginine methyl ester (l-NAME), an inhibitor of NO synthase.

Further investigation of the signaling pathway involved in insulin-mediated NO release revealed that genestein (an inhibitor of tyrosine kinase) nearly completely prevented the release of NO. Importantly, application of wortmannin, which inhibits phosphatidylinositol 3-kinase (PI3K), a signaling molecule required for insulin's effect to increase glucose uptake, caused about a 50% decrease in NO production. These in vitro results indicate that insulin induced release of NO is mediated through signaling pathways involving tyrosine kinase, PI3K, and Akt downstream from the insulin receptor (18). Importantly, Akt has recently been shown to phosphorylate endothelial NO synthase (eNos), which results in increased activity of eNos (19,20). Together, these findings suggest that insulin's metabolic and vascular actions share common signaling pathways. Thus, the similar time course of skeletal muscle vasodilation and glucose uptake in response to insulin's could be explained by a common signaling pathway. Moreover, impairment of a common signaling pathway in obesity, hypertension, or diabetes could lead to both blunting of insulin-mediated blood-flow increments and decreased rates of skeletal muscle glucose uptake. In this regard, it is important to note that mice deficient of eNOS were insulin resistant and mildly hypertensive (21), but mice deficient of endothelial insulin receptors (22) exhibited normal glucose metabolism.

Insulin's Effects on the Heart

Our lab (23) investigated the effect of different insulin infusion rates on stroke volume in groups of lean normotensive volunteers (Fig. 3A). Hyperinsulinemia in the low physi-

Fig. 3. Percent change (%A) from baseline in (A) stroke volume (SV), (B) heart rate (HR), (C) cardiac output (CO), (D) mean arterial blood pressure (MAP), and (E) total peripheral resistance (closed bar) and leg vascular resistance (hatched bar) during systemic hyperinsulinemic euglycemia and saline (control) infusion studies in lean and obese subjects. *p < 0.05, ** p < 0.01n and not significant (NS) vs baseline. (From ref. 11a.)

Fig. 3. Percent change (%A) from baseline in (A) stroke volume (SV), (B) heart rate (HR), (C) cardiac output (CO), (D) mean arterial blood pressure (MAP), and (E) total peripheral resistance (closed bar) and leg vascular resistance (hatched bar) during systemic hyperinsulinemic euglycemia and saline (control) infusion studies in lean and obese subjects. *p < 0.05, ** p < 0.01n and not significant (NS) vs baseline. (From ref. 11a.)

ological range (35 ± 4 pU/mL) and in the high physiological range (78 ± 6 pU/mL) increased stroke volume by about 7%. A nearly 15% augmentation of stroke volume was observed with supraphysiological insulin concentrations (2145 ± 324 pU/mL). A similar effect of insulin on stroke volume was reported by ter Maaten and associates (24), who observed a nearly 13% rise at insulin levels of about 50 pU/mL. The increase in stroke volume could be a result of either a decrease in peripheral resistance (see Insulin's Effect on Blood Pressure and Vascular Resistance) or as a result of an increase in inotropy of the heart muscle. Experiments in the isolated beating heart or with heart muscle preparation indicate that insulin increases contractility of heart muscle. Taken together, these data indicate that insulin has a direct effect on the heart to increase cardiac stroke volume.

In addition to augmenting stroke volume, insulin increases heart rate. In our groups, heart rate did not change at low (~35 pU/mL) levels but increased by 5% and 10% at insulin concentrations of about 80 and about 2100 pU/mL, respectively (Fig. 3B). Thus, our data indicate that insulin increases heart rate in a dose-dependent fashion. Increments in heart rate in response to hyperinsulinemia were also found by others (5,6,25) but not by all (24). The reason for the discrepancy is not clear but differences in volume status or position during the study may explain in part the different observations. Whether the increase in heart rate is a direct insulin effect or whether it is mediated by activation of the SNS is not known.

As a result of the rise in heart rate and stroke volume in response to insulin, cardiac output increases. In our study groups, cardiac output increased by about 6%, 12%, and 26% in response to insulin concentrations of about 35, 80, and 2100 pU/mL (Fig. 3C). In support of our data, Ter Maaten and colleagues found about a 9% increase in cardiac output with insulin concentrations of about 50 pU/mL (24). Moreover, Fugman and associates' study replicated most of the above findings in a more recent study (26), demonstrating increased cardiac output in response to high physiological levels of insulin. These insulin effects are not only of academic interest but may have implications under conditions in which cardiac output needs to be augmented. For example, insulin's effect to increase cardiac output has been used to improve severe heart failure in patients undergoing cardiac surgery who were unresponsive to catecholamines and vasodilators (27).

Insulin's Effects on the Sympathetic/Parasympathetic Nervous System

Insulin has been shown to increase SNSA years before its vasodilator action was appreciated (28). Systemic insulin infusion causes a dose dependent rise in NE levels. In one study (6), NE levels in response to insulin increased from 199 ± 19 pg/mL under basal conditions to 258 ± 25 and 285 ± 95 pg/mL at insulin concentrations of 72 ± 8 and 144 ± 13 pU/mL, respectively. In the same study, skeletal muscle SNSA measured by microneurograpy exhibited an even more impressive rise in response to insulin. Microneurography allows to measure frequency and amplitude of electric activity directly at the level of sympathetic nerve fibers. Determined by microneurography, SNSA increased from baseline of about 380 U to about 600 and about 750 U in response to euglycemic hyperinsulinemia. Similar differences between the methods to assess changes in SNSA have been found by others (29), suggesting that plasma NE levels may underestimate the true effect of insulin to stimulate SNSA.

Interestingly, insulin modulates SNSA in a non-uniform manner. Van De Borne and colleagues (30) studied the effect ofinsulin on skeletal muscle SNSA with microneurography. The effect of hyperinsulinemia on cardiac SNSA and parasympathetic tone was assessed by power spectral analysis of the decrease in R-R interval. Power spectral analysis allows one to distinguish between low-frequency and high-frequency components of the changes in R-R intervals. The high frequency component is thought to reflect parasympathetic nervous system activity (PNSA; vagal tone) whereas the low-frequency component reflects SNSA. Additionally, systemic infusion of the ^-blocker propranolol allows to distinguish the contribution of the PNS and the SNS on the R-R interval variability.

In response to hyperinsulinemia (84 ± 5 pU/mL), skeletal muscle SNSA increased more than twofold. In contrast, the SNSA effect on the reduction in R-R interval and variability in response to hyperinsulinemia was relatively small. This observation suggests that insulin's effect on the SNSA may be targeted specifically toward skeletal muscle the place of insulin's metabolic action. Interestingly, the increase in skeletal muscle SNSA may delay insulin's vasodilator action as proposed by Satori and associates (31).

The mechanism(s) for the increments in SNSA during hyperinsulinemia are not well understood. It may be mediated via the baroreceptor reflex to counteract insulin's vasodilator action, or may represent a direct insulin effect on the central nervous system. Moreover, coupling of insulin's effect on the SNS and its effect to increase glucose uptake/metabolism cannot be excluded. Although activation of the baroreceptor reflex in response to a decrease in blood pressure causes activation of the SNS, it can not explain all of the observed changes. First, time course of blood pressure decline and SNSA were different (6) and second, the increments in SNSA in response to insulin were nearly twice as those in response to blood pressure fall achieved by nitroglycerin infusion (32). In support of a direct role of insulin on SNSA at the level of the brain, injection of insulin directly into the third ventricle has been shown to increase SNSA in rats (33). This increase in SNSA activity could be abolished by generating a lesion in the surrounding the lateroventral portion of the third ventricle, a region implicated in the sympathetic neural control. Therefore, current evidence suggests a direct effect of insulin on the brain to increase SNSA, but other mechanisms cannot be excluded.

More recently it has been demonstrated that insulin also modulates PNSA. Unfortunately, no biochemical markers of PNSA exist that can be easily measured in vivo. As mentioned above, PNSA is studied by measuring the changes in R-R intervals using power spectral analysis. The PNSA (vagal component of heart rate control) is represented in the high frequency part of the spectrum.

In 1996, Bellavere and associates (34) reported a decrease in high-frequency variability of R-R intervals in response to hyperinsulinemia indicating that PNSA decreased. Similar results were obtained by Van De Borne and associates (30) in which euglycemic hyperinsulinemia decreased both R-R interval and the high-frequency variability of the R-R intervals. Moreover, this insulin-induced reduction of both R-R interval and high-frequency variability could not be suppressed by the ^-blocker propranolol. These data indicate that the reduction in PNSA and not increments in SNS were likely responsible for the changes in R-R interval and variability. Furthermore, these data suggest that the effect of hyperinsulinemia on cardiac SNSA may be less than originally thought.

Taken together, the current data suggest that insulin's effect to stimulate SNSA may be mediated at least in part via a direct insulin effect on the brain. Furthermore, hyperinsulinemia appears to reduce parasympathetic tone at the level of the heart, which may contribute to the increments in heart rate.

Insulin's Effects on the Kidneys

The effect of euglycemic hyperinsulinemia on renal hemodynamics has not been studied by many groups. In one study (35), insulin at levels of about 100 pU/mL has been reported to increase renal plasma flow by 10% ± 5%. A similar rise in renal plasma flow has been reported in response to L-arginine-induced insulin secretion.

Insulin's effect on electrolyte handling is well established. Insulin has been found to cause antinatriuresis (36,37), antikaliuresis, and antiuricosuria in healthy volunteers. The antinatriuresis is achieved via a decrease in fractional sodium excretion. Fractional sodium excretion fell by 20% to 30% in response to euglycemic hyperinsulinemia with insulin levels of 50 to 60 pU/mL, well in the physiological range. Reductions in potassium and uric acid excretion in response to insulin were of similar magnitude (36). Based on animal studies (38), it was thought that insulin exerts the antinatriuretic effect at the level of the distal tubule in which the highest density of insulin receptors is found but it may be that the proximal tubule is the more likely site of insulin's antinatriuretic action in humans (39). The mechanism of the antikaliuretic and antiuricoretic effects of insulin are less well elucidated.

Insulin's Effect on Blood Pressure and Vascular Resistance

Insulin's effect on skeletal muscle vasculature, stroke volume, heart rate, cardiac output, SNS, and renal sodium handling can affect blood pressure. Blood pressure is determined by cardiac output and total peripheral resistance (TPR). In other words, blood pressure in response to insulin may increase, stay unchanged or decrease dependent on the changes in cardiac output and resistance. In lean, insulin-sensitive subjects, insulin causes a small but significant fall in blood pressure. In our study (23), hyperinsulinemia in the low (35 ± 4 pU/mL) and high (72 ± 6 pU/mL) physiological range caused about a 5% drop in mean arterial pressure (MAP), and supraphsyiological insulin concentrations 2100 ± 325 pU/mL were associated with about a 10% fall in MAP (Fig. 3D). However, although a drop in MAP has been reported by many groups, it has not been observed in all studies. MAP remained unchanged in a study reported by Scherrer (7) and even increased by nearly 7 mmHg in another study (24). The reasons for the different effect of euglycemic hyperinsulinemia on blood pressure are not clear.

The decrease in MAP in light of increased cardiac output indicates (29) a fall in TPR. In fact, TPR decreased in a dose-dependent fashion by 11.1 ± 2.2, 15.0 ± 4.7, and 26.0 ± 6.0% at insulin concentrations of 35 ± 4, 72 ± 6,and 2,100 ± 325 pU/mL, respectively (Fig. 3E). A similar decrease in TPR with comparable levels of hyperinsulinemia was also observed by Fugman and associates (26). Even more impressive than the fall in TPR was the drop in leg vascular resistance (LVR). LVR decreased by nearly 45% at an insulin concentration of 35 ± 4 pU/mL (Fig. 3E). Higher prevailing insulin levels did not result in further decrements in LVR. Similar decrements resistance have been observed by Anderson in the forearm (6,40) and by Vollenweider in the calf (29). However, in one study (24) in which both blood pressure and forearm blood flow increased, no changes in vascular resistance were detected.

Metabolic Implications of Insulin's Vascular Effects

Our lab has long championed the idea that insulin's vascular effects may contribute to the rate at which glucose is taken up by skeletal muscle, which represents the majority of insulin-sensitive tissues. In other words, insulin's vascular effects may determine, at least in part, insulin sensitivity and impairment of insulin's vascular effects may result in insulin resistance.

In support of this idea, we found that insulin's effect to increase skeletal muscle blood flow and cardiac output is positively and strongly associated with the rates of glucose uptake achieved in response to euglycemic hyperinsulinemia. In two studies (23,40) performed nearly 5 years apart, the correlation coefficient between leg blood-flow increments and whole-body glucose uptake were 0.63 and 0.56, indicating that blood flow

Delicious Diabetic Recipes

Delicious Diabetic Recipes

This brilliant guide will teach you how to cook all those delicious recipes for people who have diabetes.

Get My Free Ebook

Post a comment