Cerebral energy metabolism at rest is provided almost exclusively by glucose and the molar ratio between the cerebral uptake of O₂ to that of glucose (the O₂-glucose index; OGI) is close to 6 (Quistorff et al., 2008). Adrenergic mechanisms influence cerebral energy metabolism (Bryan, 1990). Adrenaline increases the cerebral non-oxidative carbohydrate consumption (Seifert et al., 2009), presumably by a β₂ adrenergic mechanism because propranolol, a combined β₁- and β₂-adrenergic receptor antagonist, attenuates cerebral carbohydrate uptake (Schmalbruch et al., 2002; Larsen et al., 2008), whereas metoprolol, a selective β₁ adrenergic receptor antagonist, is without that effect (Dalsgaard et al., 2004). During maximal whole body exercise, the cerebral oxygen carbohydrate index (OCI; cerebral uptake of O₂/(glucose + 1/2 lactate) decreases from a resting value of ~5.7 to reach a low value of 1.7 (Volianitis et al., 2008) that is associated with high levels of plasma catecholamine (Holmqvist et al., 1986; Nielsen, 2003).

The β₂-adrenergic receptor is encoded by an intronless gene (ADRB2) located on chromosome 5 (5q31-32) that contains several single nucleotide polymorphisms (Leineweber et al., 2004). The non-synonymous 46 G > A (G16R) single nucleotide polymorphism leading to an amino acid substitution of Gly16Arg segregates with hypertension and asthma (Zaugg and Schaub, 2005; Sayers, 2013) and homozygote Gly¹⁶ subjects demonstrate a larger cardiac output (CO) both at rest and during exercise compared with homozygote Arg¹⁶ subjects (Snyder et al., 2006a; Rokamp et al., 2013). Differences in phenotype may arise from a higher receptor density in homozygote Gly¹⁶ subjects (Snyder et al., 2006b). However, differences in phenotype may also be a result of differences in sensitivity to β-agonists, as the Arg¹⁶ allele is associated with enhanced agonist-mediated desensitization (Dishy et al., 2001) and attenuated blood flow during infusion of a β-agonist in the brachial artery (Garovic et al., 2002). In contrast, Arg¹⁶ homozygotes had increased β₂-receptor sensitivity after hypoglycemia whereas no effect was seen in Gly¹⁶ homozygotes (Schouwenberg et al., 2011). Also, the G16R polymorphism has been associated with insulin resistance (Masuo et al., 2005) and obesity (Daghestani et al., 2012), albeit with inconsistent results (Gjesing et al., 2009). Another polymorphisms in the β₂-adrenergic receptor gene of functional importance is the 79C > G Q27E and in contrast to the Arg¹⁶ allele, the Glu27 allele is associated with increased agonist-mediated responsiveness in vasculature (Dishy et al., 2001). The role of haplotypes within ADRB2 is, however, not known, but Rokamp et al. (2013) found no impact of haplotypes on cardiac output.

The adrenaline driven increase in cerebral non-oxidative carbohydrate consumption (Seifert et al., 2009), if mediated by a β₂ adrenergic mechanism, could be influenced by genetic polymorphism in the β₂-adrenergic receptor. We speculated that the difference in phenotype between Gly¹⁶ homozygotes and Arg¹⁶ homozygotes could mimic that Arg¹⁶ homozygotes was influenced by a β₂ adrenergic receptor antagonist, leading to decreased cerebral carbohydrate uptake under adrenergic stress compared to Gly¹⁶ homozygotes.

No study describes the influence of genetic polymorphism in the β₂-adrenergic system on brain metabolism. We aimed to investigate cardiovascular and cerebral metabolic effects of adrenergic stimulation in humans in relation to the G16R genotype. We hypothesized that cardiac output (CO) at rest and during adrenergic stimulation would be increased in Gly¹⁶ homozygotes and that the expected reduction in cerebral metabolic ratio during adrenergic stimulation would be more pronounced in Gly¹⁶ homozygotes, reflecting increased β₂-adrenergic response compared to Arg¹⁶ homozygotes.

Forty healthy non-smoking Caucasian male subjects (age: 26 ± 5 years; height: 184 ± 6 cm; body weight: 77 ± 8 kg; body mass index: 23 ± 2 kg/m²) were included in the study following verbal and written informed consent as approved by the Comittees on Biomedical Research Ethics of the Capital Region of Denmark, The Regional Committee A (H-4-2010-027) and the Danish Data Protection Agency (2011-41-6600). To obtain groups with similar age, height, and weight the subjects were recruited from a cohort of genotyped healthy subjects (Rokamp et al., 2013). All participants completed the entire study protocol. The genotype groups included 12 G16R heterozygotes, 12 Arg¹⁶ homozygotes, and 16 Gly¹⁶ homozygotes. Age, height, weight, and body mass index were similar in the three groups (Table 1).

The subjects were studied after an overnight fast and strenuous exercise was not allowed 24 h prior to the study. Under local anesthesia (2% lidocaine), a catheter (Edwards Lifesciences, Irvine, CA) was inserted in the right internal jugular vein and advanced to its bulb using Seldinger technique. The position of the catheter was verified by a “water-fall-like” sound following infusion of saline and eventually by nociception related to the mastoid process and when so, the catheter was withdrawn about two millimeters. An arterial catheter (1.1 mm, 20 gauge) was inserted in the brachial artery of the non-dominant arm. For drug administration, a catheter (Cavafix MT134, Braun, Melsungen, Germany) was advanced to the subclavian vein through a cubital vein. Catheters were connected to a transducer (Edwards Life Sciences, Irvine, CA) positioned at heart level (5 cm below sternum) and attached to a monitor (Dialogue-2000 IBC-Danica Electronic, Denmark) for determination of mean arterial pressure (MAP) and heart rate (HR). Stroke volume (SV), CO and systemic vascular resistance (SVR) were derived by pulse contour analysis technology (BeatScope; Finapress Medical System BV, Amsterdam, Netherlands) adjusting for weight, height, age, and gender. Data were analog-digital converted and sampled at 100 Hz (Powerlab, ADInstruments, Colorado Springs, CO, USA).

Near infrared spectroscopy (INVOS-5100c, Covidien, Mansfield, MA, USA) was used to assess frontal lobe (S_cO₂) and muscle oxygenation (S_mO₂). The INVOS-5100c uses an emitter-detector distance of 3 and 4 cm and infrared light at 730 and 808 nm to avoid influence from cutaneous blood flow. One optode were applied above the supraorbital edge to assess S_cO₂ and a second optode was placed on the middle part of the thigh for assessment of S_mO₂. Transcranial Doppler sonography (2 MHz probe, Multi-Dop, DWL, Singen, Germany) determined velocity in the middle cerebral artery (MCAv_mean) from the temporal ultrasound window. The best signal-to-noise ratio was obtained at a depth of 44–56 mm with the Doppler probe secured by a headband or while handheld.

In contrast to our hypothesis, the cerebral uptake of glucose was not different among the genotype groups. Seifert et al. (2009) suggest that adrenaline is responsible for the increase in non-oxidative cerebral carbohydrate consumption. Comparing this study with the work by Seifert et al. (2009), the adrenaline infusion rate was lower (0.06 vs. 0.08 μg kg⁻¹ min⁻¹), and our subjects increased their heart rate to below 80 bpm compared to 90 bpm, suggesting a difference in adrenergic stimulation. Apart from the study of Seifert et al. (2009), the hypothesis that adrenaline should stimulate the cerebral uptake of glucose and lactate, generates from studies on the effects of exercise in rats (Schmalbruch et al., 2002) and humans (Dalsgaard et al., 2004; Larsen et al., 2008; Volianitis et al., 2008) were it is likely that the adrenergic stimulation was increased compared to that in the present study.

Snyder et al. (2006a) and Rokamp et al. (2013) found baseline CO increased in Gly¹⁶ homozygotes compared with Arg¹⁶ homozygotes. We found the same effect size (~0.5 L min⁻¹) although the difference was not statistically significant. This study differs from the previous studies by the population consisting of young males compared to mixed groups (Snyder et al., 2006a and Rokamp et al., 2013) of variable age (Rokamp et al., 2013), but a smaller population size (n = 40 compared to n = 72 (Snyder et al., 2006a) and n = 140 (Rokamp et al., 2013), respectively) and while Snyder et al. (2006a) used the open-circuit acetylene uptake method, we used the Model-flow method (Bogert and van Lieshout, 2005). Model-flow seems to underestimate the increase in CO during heat stress (Shibasaki et al., 2011), but has been successfully validated against a thermodilution estimate in healthy subject during orthostatic stress (Harms et al., 1999) and in patients with septic shock (Jellema et al., 1999), during liver transplantation and cardiac surgery (Jansen et al., 2001; Nissen et al., 2009).

The systemic increase in glucose following adrenaline stimulation is expected from increased endogen glucose production and a sustained inhibitory effect on glucose clearance (Rizza et al., 1980a). The β-adrenoceptor subtype that mediates catecholamine-induced systemic hyperglycemia is proposed to be of the β₂-subtype (Kuo et al., 1977), and the adrenaline derived decrease in glucose clearance is predominantly by a β-adrenergic mechanism (Rizza et al., 1980b). This complies with the finding that the elevation in arterial glucose is associated with the ADRB2 Gly¹⁶ polymorphism and may be explained by an elevated β₂-adrenergic response in the Gly¹⁶ homozygotes. The difference in metabolic response according to genotype may arise from a difference in number of receptors, i.e., Snyder et al. (2006b) found an increased density of β₂-receptors on lymphocytes from Gly¹⁶ homozygotes compared with Arg¹⁶ homozygotes.

In patients with longstanding type 1 diabetes, blunting of the glucagon response comes along with the disappearing endogenous insulin production (Cryer, 2008), rendering the patients increasingly dependent upon epinephrine as protection against hypoglycemia. As a result of recurrent hypoglycemia and/or long duration of diabetes, also the epinephrine response to hypoglycemia becomes blunted, leading to an increased risk of severe hypoglycemia (Høi-Hansen et al., 2010). In addition to the failing catecholamine response to hypoglycemia with impaired hypoglycemia awareness, a reduced β₂-adrenergic sensitivity has been reported in some (Korytkowski et al., 1998; Fritsche et al., 2001), but not all studies (De Galan et al., 2006). In accordance, treatment with non-specific β-blockers with effect on the β₂-receptor is associated with reduced endogenous glucose production to adrenaline infusion (Shamoon and Sherwin, 1984), impaired recovery from hypoglycemia (Lager, 1983; Popp et al., 1984), and probably an increased risk of severe hypoglycemia in type 1 diabetes. The difference between genotypes in systemic adrenergic glucose response is novel and we speculate that it may be of clinical importance in patients with type 1 diabetes. The glucose response in this study was under conditions with baseline normoglycemia and cannot be compared to the conditions during hypoglycemia. Further studies are needed to uncover the genetic impact on glucose mobilization during hypoglycemia.

The limitations of the study include not monitoring plasma insulin. There were no differences in arterial glucose level at baseline, but the Gly¹⁶ allele has been associated with increased insulin levels (Ikarashi et al., 2004) and plasma insulin increase initially in response to adrenaline (Sherwin and Saccà, 1984).

Plasma catecholamine levels are not monitored in this study, but earlier there have been found no difference between genotypes in catecholamine levels during rest or exercise (Snyder et al., 2006a).

The steady state of plasma glucose is not reached during 60 min of adrenaline infusion, and differences in plasma glucose is therefore lower than would be expected if time of adrenaline infusion had been extended.

An association was found between the G16G genotype and adrenaline induced increase in arterial glucose with no difference at baseline. We found no other relevant differences between genotypes in any other measured cardiovascular or cerebral variable at baseline or during adrenaline infusion. In conclusion, the metabolic response of glucose during adrenergic stimulation with adrenaline is associated to the G16R polymorphism of ADRB2, although without effect on cerebral metabolism.

KR participated in study design, collected the data, performed data analysis and wrote the first draft of the paper. JS, MZ, LP, and RN collected data, performed data analysis and contributed to preparation of the paper. PR performed data analysis and contributed to preparation of the paper. NS, NO, and HN participated in study design, performed data analysis, and contributed to preparation of the paper. All authors approved the manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.