INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Obstructive sleep apnea (OSA) has been associated with an increased activity of the sympathetic nervous system. Norepinephrine plasma concentration, urine catecholamine metabolites, as well as muscle sympathetic nerve activity are all increased in response to nocturnal apneic events (1). Moreover, sustained increase of these markers of sympathetic activity has been demonstrated during daytime awake rest in patients with OSA (5). Treatment of OSA by application of nasal continuous positive-airway pressure (CPAP) results in a reduction of both circulating catecholamines and muscle nerve sympathetic activity (6, 7).

Sustained increase in sympathetic activity has been demonstrated to decrease receptor density or to reduce the stimulatory response to adrenergic agonists both in animal models and in human disease (8). This phenomenon, which has been termed "receptor down-regulation" (also referred to as "refractoriness" or "desensitization"), is characterized by the diminished response to a certain constant concentration of an agonist (e.g., hormone or drug) after sustained exposure to the receptor (11). Accordingly, it has been hypothesized that the increased sympathetic activity in patients with sleep apnea induces a down-regulation of peripheral adrenergic receptors. Such down-regulation was supported by the demonstration of a diminished density and a decreased sensitivity of human B-lymphocyte ₂-receptors in patients with OSA (12). Moreover, Nelesen and coworkers (13) described reduced cardiac contractility after a mild laboratory stressor (a 3-min speaking task) in patients with OSA. Finally, Mills and coworkers (14) found a reduced cardiac chronotropic response after infusion of isoproterenol in patients with OSA.

In the present study we used the forearm vascular model to quantify the adrenergic vascular response in patients with sleep apnea and matched control subjects. The -receptor-mediated vasoconstrictory effect of norepinephrine was investigated before and after pharmacological -receptor blockade with phentolamine. In addition, the vasodilatory response to the -receptor agonist isoproterenol was evaluated.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population

All patients in this study had been referred to the sleep laboratory of Sahlgrenska University Hospital, Göteborg. This center is the main sleep laboratory for the diagnosis of sleep-related breathing disorders and the sole prescriber of CPAP units in the greater Gothenburg Health Care Area (population approximately 1,000,000). All patients had clinical signs of OSA such as snoring, witnessed apneas, and increased daytime sleepiness. Patients included in the study were males, had polysomnographically verified OSA (respiratory disturbance index [RDI]  5/h sleep), were normotensive (resting blood pressure < 140/90 mm Hg at screening without known history of hypertension), were nonsmokers, and had no active or treated coronary heart disease. Patients or control subjects with a fasting blood glucose at screening  6.7 mmol/ L or a serum cholesterol value  6.5 mmol/L were excluded from the study. Other significant medical, neurological, or psychiatric disease were excluded by medical history, physical examination, and laboratory tests. Patients did not use vasoactive drugs or other medication on a regular basis. Three patients inconsistently used dextropropoxyphen and paracetamol (analgetics, patient BL), terfenadine (antihistaminergic agent, patient PW), and ranitidine (H₂-receptor antagonist, patient CL). Any medication was stopped 2-4 wk prior to the investigation. In all, 10 patients were selected for the protocol.

Ten control subjects, roughly matched for age (± 2 yr) and BMI (± 2 kg m²), were recruited among responders to a newspaper advertisement. All control subjects had a history free of snoring and excessive daytime sleepiness. Absence of sleep apnea (RDI < 5/h of sleep) was confirmed by an overnight unattended sleep study using the Edentrace system (Nellcor/Puritan Bennett, Eden Prairie, MN). The control subjects were free of known cardiovascular and metabolic disease and any regular drug intake.

Study Procedure

Sleep studies. All patients were polygraphically assessed in the sleep laboratory prior to study inclusion. Oxygen saturation was measured with pulse oximetry (Ohmeda 3800; Datex-Ohmeda, Helsinki, Finland), oronasal airflow with thermistor, and respiratory movements with the sensitive charge bed (15) or with an impedance recording (Edentrace) (16). In addition, all patients except one were investigated polysomnographically (Embla; Flaga, Reykjavik, Iceland) within 9 mo of the experimental part of the study to confirm the diagnosis of OSA. Standard polysomnographic montage including finger pulse oximetry, oronasal thermistors, and nasal pressure measurements were used (17).

All control subjects underwent an unattended polygraphic sleep study using the Edentrace device. This system records airflow by thermistor signal, chest wall movements by means of thoracic impedance measurements (two electrode leads), cardiac rhythm, oxygen saturation by means of the finger pulse oximetry, and snoring by a laryngeal microphone. The system has previously been validated (16) and showed a sensitivity of 95% and a specificity of 96% when compared with standard polysomnography.

Only full night studies with a minimum sleep time of 5 h were accepted for assessment of diagnosis (Table 1). Recordings were scored manually using international criteria for breathing disorders (18) and sleep (19). Apnea was defined as cessation of airflow or a reduction to  20% of airflow compared with the immediately preceding baseline for at least 10 s together with a decrease of Sa_O2  4%. Hypopnea was defined as a reduction in the flow signal of at least 50% compared with the immediately preceding baseline together with an Sa_O2 reduction  4%. The RDI was expressed as the sum of apneas and hypopneas divided by the total sleep time determined by EEG analysis. In addition, RDI determined by Edentrace-recordings was calculated as the number of apneas and hypopneas in the fixed time between 12:00 A.M. and 5:00 A.M. divided by 5 h when all subjects stated to be asleep (sleep diary). Additionally, the lowest desaturation during the night was determined.

Blood flow study with forearm blood flow model. Subjects were investigated in the laboratory in the morning following a light breakfast. Caffeine-containing or alcoholic beverages were not allowed within a period of 10 h prior to the start of the investigation. Each participant was instructed to refrain from heavy exercise and to avoid emotional excitement before the experiment. Measurements took place in a quiet room with a constant temperature of 22° C and with the subject in a supine position. Forearm volume (FAV) was determined by water displacement (Archimedes principle). The amount of vasoactive drug given was adapted to the individual forearm volume by adjustment of infusion speed. An intravenous canula was placed in an antecubital vein of the dominant arm and a venous blood sample was taken for determination of plasma norepinephrine after a 30-min supine rest period. An 18-gauge polyethylene catheter (Viggo-Spectramed, Swindon, UK) was placed into the brachial artery of the nondominant arm after local anesthesia (lidocaine 1%) and connected to a Danica Dialog 2000 monitor (Danica Elektronics, Baerford, Denmark) via a DPT-6000 pressure transducer (Peter von Berg GMBH, Germany). A 30-min resting period was allowed after completion of instrumentation. Forearm blood flow was measured using a venous occlusion plethysmographic technique (20). In short, a mercury-in- silastic strain gauge was placed on the widest part of the forearm. The strain gauge was connected to an electronically calibrated plethysmograph (Elektromedicin AB, Kungsbacka, Sweden). The plethysmograph was connected to a filter pen recorder (BD 101-6; Kipp & Zonen, Delft, Holland). Venous occlusion was achieved by inflation of a cuff (80 mm Hg) placed proximal to the elbow. The hand with its arteriovenous anastomosis was excluded from the circulation by use of a second cuff, which was inflated to a suprasystolic blood pressure (approximately 200 mm Hg) starting 1 min before and maintained until the end of each blood flow measurement.

Basal blood flow was quantified during four separate measurements without infusion. Norepinephrine (NE) (Apoteksbolaget, Sweden), diluted in 5% dextrose, was then infused intraarterially at doses of 7.4, 31, 120, 472, and 1,421 pmol/100 ml FAV/min. Each infusion step was maintained for 4 min. Blood flow was measured four times during the last minute of each dosage interval. Blood pressure was recorded intraarterially immediately after each determination of blood flow.

After 30 min of subsequent rest a constant infusion of phentolamine (Rogitine; Ciba-Geigy, Stein, Switzerland) (2 µg [3.8 pmol]/100 ml FAV/min) was initiated. Phentolamine is an -adrenergic receptor antagonist acting on ₁- and ₂-receptors. Four measurements of forearm blood flow were undertaken after 10 min of phentolamine infusion. Following an identical protocol for NE infusion (doses 7.4, 31, 120, 472, and 1,421 pmol/100 ml FAV/min), the flow response to norepinephrine during concomitant -receptor blockade (phentolamine 2 µg/100 ml FAV/min) was subsequently determined.

After an additional 60 min of rest isoproterenol (Apoteksbolaget, Sweden), a -receptor agonist (dose 1, 2, 6, and 15 ng/100 ml FAV/ min corresponding to 3.8, 7.7, 23, and 58 pmol/100 ml FAV/min) was infused for 4 min at each dose step.

Plasma NE concentration samples were collected in prechilled tubes and kept on ice until centrifugation. Plasma was stored at 70° C until analysis. Samples were analyzed using a high-performance liquid chromatography technique with electrochemical detection with 3,4-dihydrobenzylamine. The lower detection limit for NE in plasma was 0.1 nmol/L. The intra- and interassay coefficients of variation were 2-4%, respectively, in the 1-2 nmol/L concentration range (21).

Data analysis. Forearm blood flow was determined by geometric analysis according to the method originally described by Whitney (22). The mean value obtained from at least four flow determinations for every measurement, each lasting a minimum of 10 s, was used for further evaluation. Forearm vascular resistance was obtained by dividing mean blood pressure (mm Hg) by forearm blood flow (ml/min/ 100 g tissue). The following data were missing. One patient did not agree to perform a polysomnography after the experimental part of the protocol due to a newly diagnosed acute leukemia. One patient and three control subjects could not tolerate the highest dose of NE. During the last part of the protocol two control subjects had invalid flow measurements at the highest isoproterenol dose due to technical failure (defect cuff inflation/pin writer). In three patients plasma catecholamines were not available due to hemolysis.

The main variables for statistical analysis were the percentage change of resistance from baseline during the NE and the isoproterenol infusions in patients and control subjects. Additional analysis was performed to compare the changes in flow, resistance, and in conductance during NE infusions with and without -receptor blockade (phentolamine) and isoproterenol infusions in patients and control subjects. For this purpose analysis of variance (ANOVA) for repeated measurements was applied (23), calculated with an SPSS 7.5 for Windows software package (Chicago, IL). The significance of the factors "dosage" (e.g., four incremental steps, NE 7.4-472 pmol/100 ml FAV/min) and "OSA" (patients versus control subjects) was tested for NE (with and without phentolamine) and isoproterenol infusions (23). To control for the differences in the baseline flow/conductance between patients and control subjects, this parameter was entered as a cofactor into the analysis. Changes in blood pressure and heart rate from baseline within the infusion protocol were tested with ANOVA including the two factors ("dosage"/"OSA") as described above. Baseline blood pressure and baseline heart rate, respectively, were entered as cofactors into the analysis.

In addition, the relation between the change in resistance with the two NE dosages 120 and 472 pmol/100 ml FAV/min (mean of z-standardized values) and resting NE concentration in plasma was determined separately in patients and control subjects using the Pearson correlation.

Baseline differences in blood pressure, heart rate, and blood chemistry between control subjects and patients with OSA were compared with Student's t-test. All tests were two tailed; a p value of 0.05 or less was considered significant and a p value of 0.05-0.1 was considered as a statistical trend. The results are shown as mean ± standard deviation (SD) in the tables and as mean ± standard error (SE) in the figures.

The study was performed in accordance with the principles outlined in the declaration of Helsinki for human research. A written consent was obtained from all participants (patients and control subjects) after oral and written information prior to the start of the investigation. The study protocol was evaluated and approved by the Regional Research Ethics Committee of Sahlgrenska University Hospital, Gothenburg.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients and control subjects did not differ significantly in terms of anthropometric measures, resting forearm vascular resistance, resting forearm blood flow, resting blood pressure, or heart rate (Tables 1 and 2). Total serum cholesterol and LDL cholesterol concentrations were slightly higher in patients compared with controls. Both plasma and urinary (24 h mean) NE concentration tended to be higher in patients compared with control subjects (2.47 ± 0.44 versus 1.94 ± 0.14, p = 0.22 and 170.8 ± 31 versus 122.3 ± 15 nmol/L, p = 0.22, respectively, Table 1), but this difference did not reach significance.

NE infusion caused a significant increase in forearm arterial resistance (p < 0.001), which was significantly attenuated in the OSA group (Figure 1, p = 0.049 for the factor "OSA"). This effect was observed for the percentage change in forearm vascular resistance (p = 0.049), the forearm blood flow (p = 0.04), and the forearm vascular resistance (p = 0.09) and the forearm vascular conductance (p = 0.04) independent from the differences at baseline (Table 2). Three control subjects, but only one patient with OSA, reported paraesthesia and pain in the forearm, hand, and/or fingers suggestive of pronounced vasoconstriction after initiation of NE infusion at the highest dose. The infusion was immediately discontinued in these four subjects and no reliable flow data were obtained at this dose. The change in resistance following the NE infusion in the high dosage range (120 and 472 pmol/100 ml FAV/min) correlated negatively with plasma NE concentrations in the OSA group (only seven patients with data on plasma catecholamines available, r = 0.76, p < 0.05) but not in control subjects (n = 10 controls, r = 0.14, p = 0.7) (Figures 2A and 2B). For the lower NE dosages 7.4 and 31 pmol/100 ml FAV/ min this correlation was negative but not significant in the OSA group. Blood pressure showed a slight, but insignificant increase during NE infusions (p = 0.18), whereas heart rate remained unchanged (Table 2).

View larger version (16K): [in this window] [in a new window]   Figure 1.   Increase in forearm vascular resistance during increasing norepinephrine (NE) infusions (dosage as log-scale, expressed as percentage reduction of forearm vascular resistance [mean ± SE]). The p value for the overall difference between OSA patients (n = 10, open circles) and control subjects (n = 10, solid circles) was 0.049 (ANOVA, between-subjects factor "OSA"). Simple contrasts between OSA patients and control subjects at separate norepinephrine dosages were not tested due to insignificant interaction between factors "OSA" and "dosage." The highest norepinephrine dose was applied in only nine patients and in seven control subjects due to clinical signs of ischemia (pain, dysesthesia).

View larger version (13K): [in this window] [in a new window]   View larger version (14K): [in this window] [in a new window]   Figure 2.   (A) The relative change in forearm vascular resistance (z-standardized mean) during norepinephrine infusion, dose 120 and 472 pmol/100 ml forearm volume/min, did not correlate significantly with resting norepinephrine plasma concentration in controls (solid circles, n = 10, r = 0.14, p > 0.5). (B) The relative change in forearm vascular resistance (z-standardized mean) during norepinephrine infusion in the high dosage range, 120 and 472 pmol/100 ml forearm volume/ min, correlated significantly with resting norepinephrine plasma concentration in patients with OSA (solid circles, only seven patients with NE plasma concentrations available, r = 0.76, p < 0.05).

Phentolamine (2 µg/100 ml FAV/min) reduced forearm vascular resistance to a similar extent in both patients and control subjects (50.8 ± 16.7% versus 43.4 ± 20.0%, respectively, p = 0.38 for factor "OSA"). Blood pressure and heart rate did not change significantly.

During ongoing phentolamine infusion patients and controls showed an almost identical response to NE (Table 3, Figure 3). Blood pressure increased slightly but insignificantly with the NE infusion (p = 0.48). Interestingly, blood pressure was significantly higher in the OSA group compared with control subjects (Table 3, p = 0.02 for factor "OSA") whereas heart rate was unchanged.

View larger version (17K): [in this window] [in a new window]   Figure 3.   Increase in forearm vascular resistance during norepinephrine infusions and ongoing phentolamine administration (2 µg/100 ml forearm volume/min) (dosage as log-scale, expressed as percentage reduction of forearm vascular resistance [mean ± SE]). The p value for the overall difference between patients with OSA (n = 10, open circles) and control subjects (n = 10, solid circles) was 0.96 (ANOVA, between-subjects factor "OSA"). Simple contrasts between OSA patients and control subjects at separate norepinephrine dosages were not tested due to insignificant interaction between factors "OSA" and "dosage."

One hour after termination of the phentolamine infusion baseline blood flow was still substantially increased, but there was no significant difference between the two groups (Table 4). Isoproterenol-induced vasodilation (1, 2, 6, and 15 ng/100 ml FAV/min) was significantly attenuated in patients compared with control subjects (percentage change in resistance, p = 0.049 for factor "OSA," Figure 4, Table 4). Blood pressure and heart rate were unchanged by the infusion protocol (Table 4).

View larger version (16K): [in this window] [in a new window]   Figure 4.   Vasodilation after isoproterenol at four different dose steps of isoproterenol (dosage as log-scale, expressed as percentage reduction of forearm vascular resistance [mean ± SE]). The p value for the overall difference between patients with OSA (n = 10, open circles) and control subjects (n = 10, solid circles) was 0.049 (ANOVA, between-subjects factor "OSA"). Simple contrasts between OSA patients and control subjects at separate isoproterenol dosages were not tested due to insignificant interaction between factors "OSA" and "dosage." The high dosage was applied only in 8 out of 10 control subjects due to technical failure.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study demonstrates an attenuated vasoconstriction after adrenergic -receptor stimulation in normotensive patients with OSA. The relative increase in resistance following NE was negatively correlated with resting plasma NE concentration in the OSA group. In addition, there was a reduced vascular response to ₂-adrenoreceptor stimulation. The results suggest a functional down-regulation of vascular - and ₂-receptors in patients with OSA.

Patients with OSA represent a group with sustained elevated sympathoadrenergic activity as evidenced by microneurographic recordings (5) as well as assessment of plasma and urinary catecholamine concentrations (24). The causative relationship between sleep-disordered breathing and sustained sympathoexcitation is supported by the reduction of sympathoadrenergic activity following treatment of the sleep apnea with nasal continuous positive airway pressure (6, 7).

The attenuated constrictor response to NE in patients with OSA may be explained by an increased activity of compensatory vasodilatory mechanisms in patients with OSA compared with control subjects. Such mechanisms could include an activation of the NO system or a concomitant ₂-receptor-mediated vasodilation. Although this possibility cannot be ruled out, there are previous data suggesting that NO-mediated vasodilation is attenuated in normotensive patients with OSA (21). Moreover, the present data demonstrated an attenuated dilatory response to isoproterenol-mediated ₂-receptor stimulation in patients with OSA making this explanation highly unlikely. It may also be argued that a structural vascular remodeling induced by OSA may have accounted for the observed effect. However, previous findings favor an increased, not a decreased, vasoconstrictor response to NE in essential hypertensives with vascular disease (25). In fact, basal blood flow was almost identical in the two groups, indirectly supporting a basically maintained resting vascular function. Also when controlling for the small difference in basal flow/conductance in the statistical analysis the attenuation of the vascular response to NE in the OSA group remained significant or showed a statistical trend. Alternatively, the attenuated vasoconstrictory response to NE may represent a down-regulation of -receptor sensitivity in peripheral vasculature in response to the sustained sympathoexcitation.

As we measured flow responses to extrinsic NE, we cannot address the question of whether a decreased number of vascular -receptors or a reduced stimulation response with a maintained number of receptors was accountable for the attenuated response. However, the negative correlation between the resting NE plasma concentrations and the vascular response to NE provides indirect support for a functional down-regulation of vascular -receptors. Finally, it is interesting to note that three controls, but only one patient, reacted with clinical signs of pronounced perfusion-limiting vasoconstriction following the highest dose of NE. This suggests that the patients with OSA better tolerated a higher circulating concentration of NE in the forearm vasculature (desensitization). Phentolamine infusion caused an approximately 50% reduction in forearm vascular resistance in both patients and control subjects, which is comparable with previous reports (26). Apparently, phentolamine induced an extensive -receptor blockade in both groups. In addition, blood pressure increased slightly in response to NE infusion in the patients whereas it decreased in control subjects. Although the reason for this difference is unclear, it may have restricted the possibility of demonstrating a difference between patients and control subjects. In fact, basal blood flow remained elevated 1 h after the phentolamine infusions, even more so in patients with OSA compared with control subjects. This may reflect a higher basal dependency on sympathetic vascular tone in the patients with OSA.

Isoproterenol-mediated vasodilation was attenuated in patients with OSA. This may have been caused by a functional down-regulation of vascular ₂-receptors similar to that observed for -receptors. Thus, our findings are in line with those of Mills and coworkers (14) who demonstrated a decreased sensitivity of cardiac ₁-receptors in patients with OSA. White and coworkers (27) demonstrated that ₁- but not ₂-receptor density decreased as a function of age. The reduced cardiac ₂-receptor responsiveness was caused by an uncoupling of receptors and a reduced G protein-mediated signal transduction. The present study population was homogenous in age but it cannot be excluded that related phenomena similar to those occurring with age may occur in OSA.

Other recent data suggest that -receptor-mediated and NO-dependent cellular cyclical GMP activation share a common mechanism for the dilatation of vascular smooth muscle (28, 29). In fact, an attenuated NO-mediated vasodilatory response has been previously demonstrated in OSA (21) and it cannot be excluded that the reduced ₂-mediated vasodilation observed in this study is at least in part due to an attenuated cellular cyclical GMP activation as a common mechanism for reduced vascular dilation. In that study Carlson and coworkers (21) suggested that an increased overall sympathetic vascular tone was one factor that might explain the observed reduced vasodilation in patients with OSA. However, the present study demonstrated that the attenuated dilatory response still persisted in the peripheral vasculature of patients with OSA independent from the concomitant -receptor blockade, which remained as a carryover after the preceding phentolamine infusion. In conclusion, there is emerging evidence for a similar pattern of functional down-regulation of cardiac ₁-, lymphocyte ₂-, and vascular ₂-receptor mechanism in patients with OSA.

The forearm vascular resistance model uses local infusion of vasoactive drugs to minimize the risk of unwanted systemic vascular effects. For instance, systemic blood pressure changes evoking a baroreflex response may have introduced an unpredictable confounder as baroreceptor function is modified in patients with OSA (30). Moreover, considerable attention was paid to carefully matching two study groups for age as ₂- receptor-mediated vasodilation and -receptor-mediated vasoconstriction both have been shown to decrease with age (31). Several studies have dealt with the forearm vascular resistance response to NE in normotensive and hypertensive subjects and this response is known to be increased in essential hypertensives (25, 26, 32). Although some of these studies have suggested a possible up-regulation of ₁- and/or ₂-receptor sensitivity as a potential pathogenic mechanism in systemic hypertension, others favor a structural change and a reduction of vascular luminal area in hypertensive subjects as an explanation for the altered resistance response. To avoid these potential confounding influences, the present study dealt with strictly normotensive subjects. Compared with control subjects, patients with OSA had slightly higher triglyceride and cholesterol concentrations, but all subjects remained within the clinically defined normal range. Although endothelium- induced vasodilation is impaired in hypercholesterolemic patients (33), recent data suggest that -receptor-mediated vasoconstriction is not influenced by elevated serum cholesterol or triglycerides (34).

The amount of sleep-disordered breathing in patients and control subjects was assessed in a first step by an ambulatory monitoring of breathing during sleep. Standard criteria for the detection of apnea and hypopnea were applied. The additional overnight polysomnography permitted a more exact determination of sleep time and breathing events in patients with OSA. Finally, NE causes vasoconstriction preferentially via -receptor stimulation. It can not be excluded that concomitant ₂-receptor-mediated vasodilation could have counterbalanced the -adrenoreceptor-mediated vasoconstriction to a different degree in the two groups. However, previous studies reported only a negligible effect on the effect of NE-induced vasoconstriction after -receptor blockade, making this a less likely explanation for the attenuated response to NE (35).

The present experiments were undertaken in normotensive patients and control subjects. It remains to be clarified whether a reduced down-regulation of -receptor function may play a role in the development of hypertension and other cardiovascular complications in some patients with OSA. Cardiovascular disease, in particular hypertension, appears to be an overrepresented, but not uniform, phenomenon in OSA (36, 37). It might be speculated that a down-regulation of -receptors provides a protective mechanism during the sympathoactivation in this disorder of sleep and breathing in some patients. Factors that may influence this protective capacity include age and genetic predisposition. Further studies of the dynamic vascular response to NE in patients with OSA with and without hypertension may shed further light on this speculation.