Insulin Attenuates Norepinephrine- Induced Venoconstriction
An Ultrasonographic Study
Abstract To directly assess insulin-related venomotor changes objectively and quantitatively, we used a modified ultrasonographic technique to measure venous diameter. Ten healthy men and women were studied by use of an Acuson 128 XP ultrasonograph with a linear 7.5-MHz ultrasonographic transducer (sensitivity, ±0.1 mm). Venous diameter was measured with the arm kept at 30° elevation and with a pneumatic cuff above the elbow inflated at 40 mm Hg for the last 2 minutes of each 5-minute observation period. Norepinephrine was infused at incremental concentrations of 12.5, 25, 50, and 100 ng/min (75, 150, 300, and 600 pmol/min, respectively) for 5 minutes each. Maximal venoconstriction was achieved by the dose of 100 ng/min norepinephrine, which was then combined with insulin doses of 8, 16, 24, and 32 μU/min (60, 120, 180, and 230 fmol/min, respectively) for 5 minutes each. In six different subjects, methylene blue, an inhibitor of guanylate cyclase, was infused simultaneously with 32 μU/min insulin and 100 ng/min norepinephrine. Mean resting diameter of the vein (1.8±0.6 mm [mean±SD]) increased (to 3.0±1.0 mm) after cuff inflation. Incremental doses of norepinephrine caused highly reproducible dose-dependent decreases in venous diameter (to 1.8±0.6 mm, P<.001). Incremental doses of insulin, when combined with the maximum dose of norepinephrine, caused highly reproducible dose-dependent increases in mean venous diameter (P<.001) compared with norepinephrine alone. Methylene blue, which had no independent effect on venous diameter, inhibited the venodilator effect of insulin (P<.05). Infusion of these substances caused no systemic changes in heart rate, blood pressure, or blood glucose. Modified venous ultrasonography can measure, in vivo, venomotor changes with hormones or drugs in doses that do not produce any systemic effects. Norepinephrine is a potent venoconstrictor whose local effects can be attenuated by insulin in normal subjects. Insulin-induced venodilation is blocked by methylene blue and is therefore probably cyclic guanylate monophosphate dependent.
The effect of vasoactive substances on the veins is an important determinant and reflection of their action on the circulation as a whole. In vitro studies on human veins have previously made use of venous strips obtained from cadavers1 and strips obtained at operation.2 Changes in venous capacitance in vivo have been measured in studies with strain-gauge plethysmography.3 Aellig4 measured change in diameter as reflected by change in electrical resistance by using a linear variable differential transformer. However, these techniques do not permit direct visualization of the vein and cannot measure the baseline or absolute diameter of the vein. Therefore, we developed a new direct visualization technique using cross-sectional and M-mode venous ultrasonography. It has the advantage of examining, in a temporal manner, direct local effects on a vein of vasoactive substances in amounts without systemic effects.
An accurate assessment of the effect of insulin on vascular tone has hitherto been difficult. First, systemic administration of insulin results in metabolic changes, such as increased glucose reuptake and use, that result in vasodilation.5 6 Second, systemic insulin can result in sympathetic stimulation that may cause vasoconstriction.7 8 Gans et al9 recently reported that insulin augmented blood pressure response to norepinephrine. Vierhapper et al10 found that hyperinsulinemia caused increases in blood pressure and aldosterone response after angiotensin II infusion in six healthy sodium-loaded men. In contrast, Yamamoto et al11 reported that larger doses of phenylephrine and angiotensin II were required to elicit pressor responses in the presence of insulin in patients with diabetes mellitus. When administered in vitro, insulin attenuates vascular reactivity to vasoconstrictor stimuli in aortic strips after endothelial denuding,12 suggesting that insulin produces a direct vasodilator effect on human vascular smooth muscle.12 13 14 Yet another recent report describes a vasorelaxant effect of insulin on veins in vivo.15 Those studies were based on the assumptions that insulin given in small doses does not cause systemic effects and that these small doses produce a highly localized effect on the smooth muscle of blood vessels infused with insulin.
The aims of our study were (1) to establish a highly sensitive reproducible method for measuring changes in venous tone, (2) to use this method to assess the vasodilator effect of insulin in physiological concentrations without systemic effects, and (3) to explore the mechanism underlying the vasomotor effects of insulin.
Fourteen healthy, normotensive volunteers aged 19 to 45 (mean±SD, 33.1±8.5 years; 10 men and 4 women) were included in the study. All were nonsmokers and had received no medication within 10 days of study. All subjects were within 20% of their ideal body weight. Ten subjects received norepinephrine and insulin infusions; 2 of these and the 4 remaining subjects received methylene blue infusions. All had normal fasting blood sugar levels (3.9 to 6.1 mmol/L), basal insulin levels (30 to 145 pmol/L), normal glucose tolerance tests, and normal lipid profiles (total cholesterol <5.2 mmol/L, low-density lipoprotein <4.9 mmol/L, and triglycerides <1.8 mmol/L).
The experiments were conducted in a room in which the ambient temperature was maintained at 26°C to 28°C. Subjects were asked to lie in a supine position and were made comfortable. The cephalic vein at the right wrist was cannulated with a 23-gauge cannula, which was heparinized, and an intravenous infusion consisting of normal saline (kept warm at 37°C) was started. Flow was maintained constant at a rate of 0.5 mL/min by a microinfusion pump (Medfusion Inc). To ensure that the volume infused was held at a constant rate (0.5 mL/min), concentration of drug(s) was altered as appropriate. The right arm was rested on a padded support elevated at an angle of 30° from the horizontal, and subjects then rested for 30 minutes. A pneumatic cuff 11 cm in width (DE Hokanson Inc) was applied 5 cm above the elbow on the right arm and connected to an inflator that elevated pressure to 40 mm Hg when required. On the contralateral arm, an automatic blood pressure cuff with heart rate monitor (Colin Medical Instruments Corp) was attached, and readings were obtained every 2 minutes to assess systemic changes.
An Acuson 128 XP ultrasonograph with a 7.5-MHz linear array transducer was used. The transducer was held stationary at a constant distance from the skin; to avoid exertion of local pressure (Fig 1⇓), a stand was specially designed. An interface of ultrasound jelly warmed to room temperature was used between the transducer and the skin. On super-VHS videocassette, a continuous recording was made of a two-dimensional image of the cross section of the vein, 1 cm distal to the tip of the cannula. A continuous recording was also made of an M-mode image, with the cursor placed at the center of the cross-sectional image on a split screen. This was used for the measurement of the diameter (Fig 2⇓).
Reproducibility of the method had been tested previously in 10 of the 14 subjects enrolled in the study. On 3 different days, repetitive measurements of a vein (site chosen in relation to a bony prominence) were taken. Videotapes were used to analyze results for testing reproducibility and intraobserver and interobserver variability.
On the infusion study day, the first two baseline readings (without inflation of the cuff and after inflation of the cuff) were taken. Norepinephrine was then infused in incremental doses of 12.5, 25, 50, and 100 ng/min (75, 150, 300, and 600 pmol/min, respectively) for 5 minutes each. During the final 2 minutes of each period, the cuff was inflated; venous measurements were taken during the last minute of each observation period. The cuff was then deflated and the next higher concentration of the drug was infused for 3 minutes with repeat of cuff inflation and measurements. After the final dose, an infusion of normal saline was resumed and the hormone drug was allowed to wash out. After 10 minutes, or when the vein had returned to its basal diameter, infusion of 100 ng/min norepinephrine was combined with insulin at incremental doses of 8, 16, 24, and 32 μU/min (60, 120, 180, and 230 fmol/min, respectively). During each 5-minute period, a constant volume of infusate (0.5 mL/min) was used, with measurements as before.
Six subjects were included in a separate experiment. This time, 100 ng/min norepinephrine and 32 μU/min insulin, the peak doses, were combined with incremental doses of 0.125, 12.5, and 125 μg/min methylene blue. In three other subjects, similar doses of methylene blue alone were infused.
Additional Control Experiments
Additional experiments were performed as further controls for the data obtained from the experiments described above: (1) Insulin vehicle (0.25% phenylmethyl cresol) alone in place of insulin was used, and the remainder of the protocol was not altered; (2) in six subjects, insulin alone in the doses of 8 to 32 μU/min was infused without norepinephrine; (3) in three subjects, the maximum dose of norepinephrine, 100 ng/min, was infused in four cycles of 5 minutes each after the initial doses of 12.5 to 100 ng/min norepinephrine as described above (no insulin was infused in this experiment), and (4) in three subjects, methylene blue (0.125, 12.5, and 125 μg/min) was combined with 100 ng/min norepinephrine after initial doses of 12.5 to 100 ng/min norepinephrine as described above.
In two subjects, plasma insulin and glucose levels were determined by samples taken from the control contralateral arm at 2.5 minutes (without the cuff inflation) and 4.5 minutes (after the cuff inflation) into each dose of insulin infusion.
All diameters are expressed in millimeters as mean±SD. Basal diameter of each vein was measured during saline infusion before and after ipsilateral cuff inflation to 40 mm Hg. Postinflation diameters were used in all subsequent comparisons (Table⇓). Comparison between diameters obtained during norepinephrine infusion and combined insulin and norepinephrine infusions were made with a repeated-measures ANOVA using the systat program. One-way ANOVA was used to assess the results of the low-dose methylene blue infusion.
Validity and Reproducibility
The maximal variability between two measurements of the same recording by the same observer on four different occasions was 0.1 mm. The maximal difference between four measurements of the same recording by a second observer in the recordings of all 10 subjects in whom the reproducibility studies were carried out was 0.1 mm. Thus, the interobserver coefficient of variation of the technique was less than 5% for veins of 2.0 mm diameter. There was no time-dependent change in venous diameter when normal saline (0.5 mL/min) was infused for 30 minutes in 4 different subjects. The measured diameter of the vein varied by less than 0.1 mm under these conditions for the 10 measurements carried out in each of the subjects. Thus, the time-related within-subjects coefficient of variation was also less than 5%.
Effect of Norepinephrine on Venous Diameter
The mean baseline diameter of the vein was 1.8±0.6 mm (range, 0.9 to 2.6 mm). It increased on inflation of the cuff (at 40 mm Hg) to 3.0±1.8 mm (range, 1.8 to 5.1 mm). This value is referred to as the basal (inflated) diameter on normal saline (Table⇑). On administration of 12.5 ng/min norepinephrine, the smallest dose, the venous diameter decreased in 4 subjects (by 13% to 26%); mean venous diameter for the group of 10 was 2.9±1.0 mm (range, 1.4 to 5 mm). The dose of 100 ng/min decreased the diameter in all 10 subjects, the mean diameter being 1.8±0.6 mm (range, 0.7 to 2.6 mm; Fig 3⇓). The results were highly significant (F=14.8, P<.001 for within-subjects variation; F=44, P<.001 for linear test of order effects). After a 10- to 12-minute washout period, while saline was being infused, venous diameters returned to baseline (1.6±0.1 mm without cuff inflation and 3.1±0.9 mm after cuff inflation). The second phase of the experiment was then started.
Effect of Insulin on Preconstricted Veins
The norepinephrine infusion rate of 100 ng/min caused consistent optimum constriction in all subjects. This dose of norepinephrine was then continuously infused in combination with incremental doses of insulin at 8, 16, 24, and 32 μU/min. The comparison of mean venous diameters achieved with insulin is shown in the Table⇑. The dose-dependent increase in venous diameter caused by insulin in veins preconstricted by peak dose norepinephrine was highly significant (F=17.6, P<.001 for within-subjects variation; F=36, P<.001 for linear test of order effects; Fig 4⇓). When insulin at 32 μU/min was combined with 100 ng/min norepinephrine, the mean venous diameter achieved was 2.9±1.0 mm (range, 1.6 to 4.6 mm).
Effect of Methylene Blue in Combination With Insulin and Norepinephrine on Venous Diameter
Six subjects completed all or part of the methylene blue infusions. In this group, vein diameter during peak dose norepinephrine infusion (100 ng/min) was 1.4±0.3 mm. Addition of 32 μU/min insulin to norepinephrine increased mean venous diameter to 2.1±0.4 mm. Subsequent addition of methylene blue at 0.125 μg/min (to norepinephrine plus insulin) caused venous diameter to return to baseline (1.3±0.3 mm) (Fig 5⇓). Higher doses of methylene blue exerted no further effects in attenuation of insulin actions. In three subjects’ forearms, pain was experienced with the higher doses of methylene blue despite absence of other local or systemic responses.
Additional Control Experiments
In separate experiments, the following findings were made: (1) Insulin vehicle (0.25% phenylmethyl cresol) infused with norepinephrine did not have an effect on venous diameter; (2) insulin alone did not have a venodilator effect at these infusion rates; (3) infusion of norepinephrine alone in the dose of 100 ng/min in four cycles of 5 minutes each, after the initial four cycles of norepinephrine described above, did not produce any loss of vasoconstrictor effect (tachyphylaxis); and (4) infusion of methylene blue after norepinephrine infusion did not further constrict the vein. During infusions of norepinephrine and insulin, blood pressure, heart rate, plasma glucose, and symptoms were monitored. No change was noted in any of these indexes during infusion. Plasma glucose and insulin concentrations measured in blood samples obtained 10 cm downstream from the site of infusion and in those obtained from the contralateral arm were not altered during any of the infusions.
Our data demonstrate that ultrasonography of the cephalic vein at the wrist can provide consistent and reproducible results when used to test vasoactive substances. The exquisite sensitivity and the precision of our system, the appearance of the vasoconstrictor effect within 6 seconds of the drug’s reaching the vein, and the sensitivity (±0.1 mm of change) allow physiological experimentation. Using this technique, we can demonstrate constriction with 25 ng/min norepinephrine; constriction of approximately 50% was observed with 100 ng/min norepinephrine without any systemic effects. These data are comparable to those of Martin et al,16 who demonstrated in 58 subjects of both sexes (age range, 16 to 78 years) that the dose of norepinephrine required to produce 50% constriction varied from 1.5 to 300 ng/min (10 to 1800 pmol/min) and was unrelated to the age of the individual. Measurements were carried out with the cuff inflated at 40 mm Hg because without cuff inflation, increased variability and spontaneous pulsatility of the vein appeared. We observed variability of up to 0.3 mm in the same vein without inflation when measurements were taken at different times. This reflects spontaneous pulsatility rather than technical variability. Our method allows a direct measurement of venous diameter and thus permits direct comments on constriction and dilation rather than derivation from observation on pressure and tension. The coefficient of variation of the technique was less than 5%.
An important observation in this series of experiments is that insulin induces a dose-dependent inhibition of the vasoconstrictor effect of norepinephrine. We were unable to demonstrate a direct vasodilator effect of insulin. Our experience also supports earlier observations in the literature17 that in subjects who are supine and comfortably relaxed, the forearm veins are usually fully dilated, in part because of low sympathetic output. In these circumstances no response is seen if a dilator substance is given. The observation that insulin alone does not cause vasodilation is compatible with the view that control of venous diameter rests on a balance between the constrictor and the dilator forces. In the present experiment, that balance was further altered by the experimental design. The markedly diminished sympathetic output and the increase in the pressure in the veins by cuff inflation to 40 mm Hg tipped the balance in favor of maximal venous distension, which would reduce the ability to observe the venodilator effect of any substance, including insulin.
The infusion rates at which we observed vasoactive effects of insulin were less than 32 μU/min, with a threshold effect at 8 μU/min in some of the subjects. This suggests that insulin exerts its effects within the physiological range. Because neither norepinephrine nor insulin induced a systemic effect on heart rate, blood pressure, or blood glucose concentrations, and because the effects appeared within seconds of administration of norepinephrine or insulin, the observed effects are probably local. Present data are qualitatively similar to those obtained by Feldman and Bierbrier15 on dorsal hand veins and by Creager et al18 on human brachial arteries. Creager et al concluded that insulin causes forearm vasodilation primarily by a β-adrenergic mechanism. At the higher doses used in their study, insulin may have decreased forearm vascular resistance directly or by a mechanism other than adrenergic stimulation. Feldman and Bierbrier infused insulin at rates between 1 and 300 mU/min (7 and 2160 fmol/min), the dose producing 50% of the maximum effect being 4.2 mU/min (30 fmol/min). These doses were greater than ours by several orders of magnitude. We observed 50% of the maximum insulin effect at 16 μU/min. At the highest rates of infusion used in our experiments, the maximum possible concentration achieved at the infusion site would be expected to be 64 μU/mL (460 fmol/min). The highest dose administered was 32 μU/min in 0.5 mL saline; during the venous stasis phase at 40 mm Hg, the infused insulin is not expected to flow in an antegrade fashion. The fasting concentration of insulin in normal subjects in our laboratory is <20 μU/mL (<145 pmol/L), and concentrations after a 75-g glucose challenge increase to between 40 and 80 μU/mL (290 and 575 pmol/L).
Some authors have suggested that insulin stimulates Na+-H+ exchange and also stimulates the sodium-potassium pump, which leads to hyperpolarization of vascular smooth muscle, diminished calcium influx, and relaxation.19 Others have shown diverse effects in strips of rabbit femoral artery20 or human forearm.21 The mechanism proposed by Zemel and coworkers22 is that insulin may promote vascular relaxation and inhibition of vasoconstrictive responses by causing increased Ca2+-ATPase activity in the plasma membrane, thereby increasing calcium efflux from vascular smooth muscle.
Our data also provide a clue to the mechanism underlying the venous effects of insulin. If insulin exerts its effect by activating either nitric oxide synthase or guanylate cyclase, it would increase levels of cyclic guanosine monophosphate (cGMP). Methylene blue, a known inhibitor of guanylate cyclase and nitric oxide synthase, inhibited the vasodilator effect of insulin. It therefore seems likely that insulin’s venodilator effects are cGMP dependent. cGMP is a known vasodilator and platelet antiaggregator. Insulin’s antiaggregating effect on platelets is dependent on guanylate cyclase activation, in keeping with present observations.23 Our data do not, however, demonstrate whether the effect of insulin on guanylate cyclase is a direct one or whether it is mediated by the generation of nitric oxide. This issue is currently under investigation in our laboratory. It is noteworthy that two groups of investigators have recently demonstrated that the vasodilator effect of insulin on the arterial side is nitric oxide mediated.24 25 Their conclusions were based on plethysmographically measured increases in flow in the forearm25 and increases in femoral artery blood flow measured by a thermodilution catheter24 rather than the measurement of the arterial diameter. Measurements of blood flow do not necessarily reflect changes in vessel diameter, because this increase can be a reflection of vasodilation in the microvasculature and increased blood velocity in the macrovasculature.
Insulin resistance is known to be associated with hypertension and an increased risk of vascular disease. Hitherto, data on the mechanisms linking insulin resistance with vascular disease have been scanty and the understanding minimal. The demonstration of impaired vascular responses to insulin in insulin-resistant patients should provide a bridge between insulin resistance, vascular diseases, and hypertension. Indeed, Feldman and Bierbrier15 have demonstrated recently that the effects of insulin on dorsal veins of the hand are impaired in obese and hypertensive persons. Both obesity and hypertension are known to be associated with insulin resistance.
In conclusion, venous ultrasonography is a sensitive and reproducible method for measuring venous diameter and testing vasoactive substances. In our experiments, we used it to describe the cGMP-dependent vasodilator effect of insulin; this effect may have important metabolic and clinical implications.
The authors acknowledge the help of John C. Makowski, MLT (ASCP), Lisa M. Martin, RN, and Mary K. Bateson, RN, in conducting the study and of John Love and Kuldip Thusu in helping with the laboratory results. Appreciation is also expressed to Jacqueline Blackley, RN, and Arthur Orlick, MD, for their valuable work in obtaining the research Acuson 128 XP instrument.
Maloff G. Pharmakologische Versuche an isolierten Venen des Menschen. Pflugers Arch Physiol. 1942;229:38-42.
Coupar IM. The effect of isoprenaline on adrenoreceptors in saphenous vein. Br J Pharmacol. 1970;13:465-475.
Wood EJ, Eckstein JW. A tandem forearm plethysmography for study of acute responses of the peripheral veins of man: the effect of environmental and local temperature change and the effect of pooling blood in the extremities. J Clin Invest. 1958;37:41-50.
Baron AD, Bretchtel-Hook G, Johnson A, Hardin D. Skeletal muscle blood flow: a possible link with insulin resistance and blood pressure. Hypertension. 1993;21:129-135.
Anderson EA, Mark AL. The vasodilator action of insulin: implication for the insulin hypothesis of hypertension. Hypertension. 1993;21:136-141.
Gans RO, Donker AJ. Insulin and blood pressure regulation. J Intern Med. 1991;73(suppl):49-64.
Gans RO, Bilo HJ, von Maarschalkerweerd WW, Heine RJ, Nauta JJ, Donker AJ. Exogenous insulin augments in healthy volunteers cardiovascular reactivity to noradrenaline but not to angiotensin II. J Clin Invest. 1991;88:512-518.
Zemel MB, Reddy S, Sowers JR. Insulin attenuation of vasoconstrictor responses to phenylephrine in Zucker lean and obese rats. Am J Hypertens. 1991;4:537-539.
Kim YC, Zemel MB. Insulin increases vascular smooth muscle recovery from intracellular calcium loads. Hypertension. 1993;22:74-77.
Kahn AM, Siedel CL, Allen JC, O’Neil RG, Shelat H, Song T. Insulin reduces contraction and intracellular calcium concentration in vascular smooth muscle. Hypertension. 1993;22:735-742.
Creager MA, Liang CS, Coffman JD. Beta adrenergic-mediated vasodilator response to insulin in the human forearm. J Pharmacol Exp Ther. 1985;235:709-714.
Ferrannini E, Taddei S, Santoro D, Natali A, Boni C, Del Chiaro D, Buzzigoli G. Independent stimulation of glucose metabolism and Na+-K+ exchange by insulin in the human forearm. Am J Physiol. 1988;255:E953-E958.
Yagi S, Takata S, Kiyokawa H, Yamamoto M, Noto Y, Ikeda T, Hattori N. Effect of insulin on vasoconstrictive response to norepinephrine and angiotensin II in rabbit femoral artery and vein. Diabetes. 1988;37:1064-1067.
Sakai K, Imaizumi T, Masaki H, Takeshita A. Intra-arterial infusion of insulin attenuates vasoreactivity in the human forearm. Hypertension. 1993;22:67-73.
Zemel MB, Johnson BA, Ambrozy SA. Insulin-stimulated vascular relaxation: role of Ca2+-ATPase. Am J Hypertens. 1992;5:637-641.
Trovati M, Massucco P, Mattiello, Mularoni E, Cavalot F, Anfossi G. Insulin increases guanosine-3′,5′-cyclic monophosphate in human platelets: a mechanism involved in the insulin anti-aggregating effect. Diabetes. 1994;43:1015-1019.
Steinberg HO, Brechtel G, Johnson A, Fineberg N, Baron A. Insulin-mediated skeletal muscle vasodilation is nitric oxide dependent: a novel action of insulin to increase nitric oxide release. J Clin Invest. 1994;94:1172-1179.
Scherrer U, Randin D, Vollenweider P, Vollenweider L, Nicod P. Nitric oxide release accounts for insulin’s vascular effect in humans. J Clin Invest. 1994;94:2511-2515.