Limited Acute Influences of Electrical Baroreceptor Activation on Insulin Sensitivity and Glucose Delivery: A Randomized, Double-Blind, Crossover Clinical Study

Insulin-mediated glucose uptake in skeletal muscle depends on glucose delivery to the interstitial space via skeletal muscle blood vessels. Arterial baroreceptors regulate sympathetic nerve traffic to muscular resistance vessels (1) and may thus indirectly affect muscular glucose and insulin delivery. Indeed, disorders associated with impaired baroreflex function including type 2 diabetes, arterial hypertension, and chronic heart failure are accompanied by excessive sympathetic activity as well as insulin resistance (2). Chronic sympathetic inhibition with moxonidine improved insulin sensitivity in insulin-resistant, obese hypertensive patients (3). In contrast, acute sympathetic activation decreased forearm blood flow and insulin-mediated glucose uptake in healthy human volunteers (4). Remarkably, reflex-mediated vasoconstriction was more effective than norepinephrine-induced vasoconstriction in decreasing insulin sensitivity in this experimental setting (5). Patients with treatment-resistant arterial hypertension who have been implanted with electrical carotid sinus stimulators (6) provide the unique opportunity to study acute baroreflex influences on human metabolism. We previously showed that baroreceptor activation through electrical carotid sinus stimulation decreases efferent muscle sympathetic nerve activity and blood pressure within seconds (7). We now tested the hypothesis that acute changes in electrical carotid sinus stimulation alter insulin sensitivity by changing muscular glucose delivery.

Treatment-resistant hypertensive patients who had been implanted with an electrical carotid sinus stimulator at least 6 months before study initiation were eligible to participate in this study. Patients had to have an acute reduction in systolic blood pressure of >10 mmHg during electrical dose-response testing on days 2–4 after carotid sinus stimulator implantation surgery. Before electrical carotid sinus stimulator implantation, patients had to have an office systolic blood pressure >160 mmHg or diastolic blood pressure >90 mmHg despite treatment with three or more antihypertensive medications at full doses, including a diuretic. Secondary causes of arterial hypertension had been excluded in all patients. The Hannover Medical School ethics committee approved the study, and all patients gave written informed consent.

Patients underwent metabolic testing twice after an overnight fast. We tested patients with active electrical stimulation or the stimulator switched off in a randomized, double-blind, and crossover study on separate days. The stimulator was switched off 2 h before insulin sensitivity measurements started or was left on according to the randomization list by a physician not involved in metabolic tests. For ethics and safety reasons, patients were studied on their chronic antihypertensive medication. Patients with diabetes were allowed to take oral antidiabetic medication but not to inject antidiabetic drugs before and during the tests. Drug treatment was not changed throughout the study. One hand was instrumented with a venous catheter and placed in a heating pad to allow perfusion of arterio-venous anastomoses for the collection of arterialized blood samples. We inserted another venous catheter on the contralateral arm for glucose and insulin infusion. A microdialysis catheter (CMA 60; CMA/µDialysis, Stockholm, Sweden) was inserted into the vastus lateralis muscle as previously described (8,9). The microdialysis catheter was perfused with lactate-free Ringer solution (CMA/µDialysis) supplemented with 30 mmol/L ethanol solution at a flow rate of 2 µL/min. Catheter equilibration was allowed for at least 60 min before baseline measurements started.

Two hours after true or sham manipulation of the stimulator, we submitted patients to a frequently sampled intravenous glucose tolerance test (FSIGT) with a 30-min baseline period, glucose injection at minute 0, and insulin injection at minute 20 (10). Arterialized blood samples were obtained over 180 min to measure glucose and insulin. We calculated areas under the curve and applied the Minimal Model Assessment technique to calculate glucose effectiveness, the acute insulin response to glucose, insulin sensitivity, and the disposition index (11). Microdialysis samples were taken every 15 min and were analyzed for glucose, lactate, pyruvate, and urea using the automated enzyme-linked spectrophotometric CMA 600 analyzer (CMA/µDialysis). Ethanol in perfusates and in microdialysates was determined as previously described (12). Ethanol dilution and microdialysate urea concentrations served to assess changes in tissue blood flow (13,14).

Data are given as mean ± SEM unless otherwise stated. Parameters with the stimulator on or off were compared by a two-sided t test using IBM SPSS Statistic 20. A value of P < 0.05 was considered significant. The primary end point was the difference in insulin sensitivity with the carotid sinus stimulator switched on and off.

We contacted 25 patients; 5 refused to participate, and 4 were not eligible after consenting. Sixteen patients underwent complete insulin sensitivity testing. Patients took on average six different antihypertensive medications, and mean duration of hypertension was 7.5 years (range 4–11). Figure 1 illustrates maximal blood pressure and heart rate reductions (mean ± SD: −45 ± 5/−23 ± 4 mmHg and −9 ± 2 bpm) during dose-response tests 1 month after surgery. Baseline characteristics of the study population are summarized in Table 1. Two implantable electrical carotid sinus stimulators are in use (all by CVRx Inc., Minneapolis, MN) (15). The Rheos device (12 patients) consists of two carotid sinus electrodes, which are bilaterally implanted and provide electrical stimulation via five finger-like electrodes attached around the carotid sinus. The Neo device consists of one carotid sinus electrode. This system was used in four patients, but a second Neo device was implanted for bilateral stimulation in two patients. Three patients also had earlier catheter-based renal nerve. Mean supine blood pressure 2 h after stimulator manipulation was 151/88 (“on”) and 164/91 (“off”; P < 0.05 for systolic blood pressure).

Six patients presented with a fasting glucose concentration >7.0 mmol/L and were on antidiabetic drugs. Mean duration of diabetes was 3 years (range 1–4). Fasting homeostasis model assessment index of insulin resistance in patients without diabetes ranged between 0.7 and 3.5 with 8 of 10 patients fulfilling the criterion for insulin resistance (homeostasis model assessment >2.5). Before and throughout FSIGT, mean arterialized glucose and insulin concentrations were virtually identical with the stimulator switched on or acutely switched off (Fig. 2). The area under the curve of glucose (30 min baseline up to 180 min after glucose injection) was 1,470 ± 102 mmol/L ⋅ min with the stimulator on and 1,534 ± 119 mmol/L ⋅ min with the stimulator off (P = 0.27). Insulin sensitivity, the primary end point of this study, was 3.3 ± 1.0 (mU/L)⁻¹ ⋅ min⁻¹ with the stimulator on and 4.4 ± 2.6 (mU/L)⁻¹ ⋅ min⁻¹ with the stimulator off (P = 0.7). Glucose effectiveness was 0.019 ± 0.003 (on) vs. 0.018 ± 0.002 min⁻¹ (off; P = 0.5), disposition index was 603 ± 145 (on) vs. 949 ± 501 (off; P = 0.5), and acute insulin response to glucose was 190 ± 50 (on) vs. 181 ± 40 mU ⋅ L⁻¹ ⋅ min (off; P = 0.71).

We obtained complete microdialysis sample sets in 13 of 16 patients owing to medical and technical reasons. Interstitial urea concentrations and the ethanol ratio were similar with the stimulator on and with the stimulator off throughout FSIGT, demonstrating comparable blood flow in the tissue surrounding the microdialysis catheter. Figure 3 illustrates microdialysis metabolite concentrations during FSIGT and individual area under the curve data for each metabolite. Interstitial concentrations of glucose, lactate, and pyruvate increased during FSIGT, but the response was similar with electrical carotid sinus stimulation on or off.

The main finding of our study is that an acute change of chronic electrical carotid sinus stimulation in patients with treatment-resistant hypertension does not elicit significant changes in muscular glucose delivery and systemic insulin sensitivity. Because carotid sinus stimulators provide for the first time the possibility to experimentally manipulate baroreceptor input for more than a few seconds in patients (15), our findings provide novel insight into the contribution of neural mechanisms to the regulation of insulin and glucose metabolism.

The importance of baroreflex mechanisms in short-term regulation of sympathetic activity and vascular tone is undisputed. Observations in sinoaortic denervated dogs suggested that baroreflexes do not control arterial blood pressure in the long-term (16), but in more recent investigations, partial baroreceptor denervation combined with baroreceptor unloading produced a sustained blood pressure increase (17). Moreover, electrical carotid sinus stimulation elicited chronic blood pressure reduction in animal models and in patients with treatment-resistant arterial hypertension (6,18). The finding that baroreflexes contribute to long-term control of sympathetic activity and vascular tone in dogs provided an impetus for conducting the present investigations (19).

A significant proportion of patients implanted with electrical carotid sinus stimulators did not respond with blood pressure reduction during acute stimulation in our previous study (7). We only included responders defined as patients showing an acute reduction in systolic blood pressure of at least 10 mmHg with electrical stimulation during the first days after surgery. Thus, the patient population included in our study was suitable to assess influences of electrical carotid sinus stimulation on glucose and insulin metabolism.

Muscular insulin sensitivity is determined by vascular glucose and insulin delivery to the muscular interstitial space, insulin-mediated cellular glucose uptake, and intracellular glucose metabolism (20). According to earlier studies (4,5), we expected to observe decreased muscular glucose delivery and insulin-mediated glucose disposition and metabolism when electrical carotid sinus stimulation was acutely interrupted. Yet, patients showed an identical metabolic response during FSIGT with the stimulator switched on and with the stimulator switched off. Apparently, insulin sensitivity responds to changes in sympathetic activity in a heterogeneous way. Intrabrachial infusion of the nonselective α-adrenoreceptor antagonist phentolamine acutely augmented forearm glucose uptake in congestive heart failure patients but not in healthy young individuals (21). Moreover, axillary plexus blockade increased forearm blood flow but did not improve glucose uptake in healthy individuals (22). Influences of sympathetic activity on insulin sensitivity may thus be restricted to patients with comorbid conditions, such as heart failure. Remarkably, insulin sensitivity did not respond to baroreflex-mediated changes in sympathetic tone even though our patients were older, mostly insulin resistant, and severely hypertensive.

One limitation of our study is that for safety reasons, we limited the duration of periods without electrical baroreflex stimulation to a few hours. We cannot exclude that the duration of the change in electrical carotid sinus stimulation was too short to affect glucose metabolism. For ethical reasons, we had to study our patients on various medications, which could have masked an effect of electrical carotid sinus stimulation on blood flow and glucose metabolism. For example, vasodilator treatment may attenuate vascular responses to changes in electrical baroreflex stimulation. Finally, patient characteristics or experimental conditions may also have contributed to our findings. Presence of long-standing treatment-resistant hypertension and concomitant vascular disease and the fact that we conducted measurements in the supine position, which limits basal sympathetic activation, could have affected the metabolic response.

Yet, carefully conducted physiological studies in animals support our findings. In a dog model of diet-induced obesity, arterial hypertension, and insulin resistance, electrical baroreflex stimulation for 1 week elicited a sustained reduction in sympathetic activity and blood pressure. However, electrical baroreflex stimulation did not ameliorate insulin resistance in obese dogs (23).

Another important limitation of our study is the relatively small number of patients. We had to recruit patients from several German sites. Based on the average insulin sensitivity of 3.3 ± 1.0 (mU/L)⁻¹ ⋅ min⁻¹ with the stimulator on, our study had an 80% statistical power to detect an effect size of 0.75 with n = 16 and α = 0.05. In fact, the true effect size in our study was only 0.18. Thus, patient number does not interfere with our conclusion. Despite these limitations, we propose that acute baroreflex-mediated changes in resistance vessels constriction have a limited effect on resting muscular glucose delivery and whole-body insulin sensitivity in patients with severe treatment-resistant arterial hypertension.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. M.M. wrote the manuscript and researched data. J.A. researched data. J.M., H.H., J.B., and S.Ec. reviewed and edited the manuscript. J.J. wrote the manuscript. S.En. wrote the manuscript and researched data. S.En. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.