Microvascular oscillations are influenced by central and local factors. Although oscillations of local origin, also termed “vasomotion,” have been observed since the invention of the microscope (1), their physiological role and underlying mechanism are not fully understood (2). Anesthetics and sedative drugs have a major influence on the macro- and microcirculation, and the effects of reduced sympathetic nervous activity are well-documented. However, results from in vivo and in vitro studies on the effect on smooth vascular muscle and endothelium are conflicting. Since vasomotion may improve tissue oxygenation, especially in conditions with limited perfusion (3), knowledge of how anesthetic drugs influence vasomotion should be of clinical interest.
Changes in human skin microcirculation, including vasomotion, can be evaluated noninvasively by analyzing laser Doppler flowmetry signals using wavelet transform (4). In the frequency interval 0.0095–2 Hz, oscillations at five different frequencies can be identified, including centrally mediated oscillations from the heart, ventilation, and the sympathetic nervous system. Data indicate that the frequencies around 0.1 Hz are related to local mechanisms, such as intrinsic myogenic activity of vascular smooth muscle cells (5,6). The slowest waves around 0.01 Hz can be modulated by acetylcholine (ACh), indicating a connection to the vascular endothelium (7).
We hypothesized that wavelet transform of laser Doppler flowmetry signals can detect changes in the microcirculation induced by general anesthesia. To our knowledge, there are no reports on these alterations in human skin during general anesthesia. In the present study, we compared oscillations in human skin microcirculation before and during general anesthesia in a group of patients undergoing faciomaxillary surgery. Eleven healthy volunteers served as a control group.
The study was approved by the local ethics committee of Oslo. Eleven healthy patients undergoing faciomaxillary surgery (18–30 yr) were included. Written informed consent was obtained from all patients. All were nonsmokers not taking regular medication and were ASA I classified. Eleven healthy, nonsmokers were also included as a control group. One of these subjects was excluded due to a stress reaction in one of the recordings. Anthropometric data are reported in Table 1.
Recordings were performed with subjects in a supine position in a quiet room. The room temperature was kept at 23°C. The experimental procedure started after an acclimatization period of 20 min. Two probes for basal laser Doppler flowmetry recordings (MP1 probes, Moor Instruments, Axminster, Devon, UK) and two iontophoretic chambers combined with laser Doppler flowmetry probes (DP1T probes, Moor Instruments) were positioned on the volar aspect of the lower arm. All probes were located on the same arm, at least 5 cm apart, avoiding superficial veins and damaged skin areas. The probes for basal laser Doppler flowmetry recordings were kept in the same positions during the experiment. To avoid residual effects of the drugs, the iontophoretic chambers were moved to untreated skin areas after the first recording. Recordings for an interval of 22 min were obtained before and during general anesthesia. The second recording during general anesthesia started approximately 15 min after induction of anesthesia. The average time between the two registrations was 50 min. Arterial blood pressure, ventilation frequency, and heart rate were measured during both registration periods. In the control group, there were two separate laser Doppler flowmetry registrations without any interventions between. The protocol was similar except that anesthesia was not induced. The time interval between the two registrations was 20 min.
General anesthesia was induced by a target-controlled propofol infusion (5–7 μg/mL), fentanyl (2 μg/kg), and midazolam (20 μg/kg). A laryngeal mask was inserted when the jaw was relaxed, and the patient was then mechanically ventilated in a pressure-controlled mode, giving an end-tidal Pco2 range within 4.7–6.0 kPa (Pneupack 770, System 2, UK). The ventilation rate was set to 12 breaths/min. General anesthesia was maintained with propofol-target-controlled infusion adjusted to 2–3 μg/mL. No IV fluids or additional medication to increase arterial blood pressure were given. No complications were registered.
Laser Doppler Flowmetry
Laser Doppler flowmetry gives a semiquantitative measurement of microvascular blood perfusion, expressed in arbitrary units (AU) (8). Laser Doppler flowmetry measurements from the skin reflect perfusion in capillaries, arterioles, venules, and the dermal vascular plexa. A major part of the signal reflects thermoregulatory perfusion (9). The Laser Doppler flowmetry measurements were obtained with a two-channel flowmeter (MoorLAB server/ satellite, Moor Instruments, Axminster, Devon, UK) for basal, unstimulated recordings (probes 1 and 2) and a two-channel flowmeter (DRT 4, Moor Instruments, Axminster, Devon, UK) for recordings during iontophoresis with ACh and sodium nitroprusside (SNP) (probes 3 and 4). A sampling frequency of 40 Hz and a time constant of 0.1 s were selected.
Iontophoresis allows transdermal delivery of polar drugs by means of a small electrical current. It is then possible to assess microvascular reactivity, using laser Doppler flowmetry to measure skin perfusion (10). Two combined probe holders for iontophoresis and perfusion measurement (MIC1-E ION, Moor Instruments, Axminster, Devon, UK) of opposite polarity were attached to the volar side of the forearm and connected to a constant current stimulator (MIC 1, Moor Instruments, Axminster, Devon, UK). A solution of 1% ACh and 1% SNP were delivered separately to each iontophoresis chamber to induce endothelium-dependent and -independent vasodilatation. Charges of 0.75 mC (75 mA for 10 s), 1.5 mC (150 mA for 10 s), 3.0 mC (150 mA for 20 s), and 6.0 mC (200 mA for 30 s) were applied. After each charge, there was a measuring period of 300 s. The statistical analysis is based on the mean value of the total period of recording.
Spectral Analysis and Wavelet Transform
Laser Doppler signal from human skin are nonlinear. Like heart-rate variability, the characteristic frequencies of the laser Doppler signal continuously change with time. The oscillations span a wide frequency range. This implies the necessity for good low frequency resolution and good time resolution for higher frequency components. The continuous wavelet transform, based on Morlet mother wavelet, is performed to achieve optimal localization in both time and frequency domains (11,12). With an adjustable window length, an optimal frequency resolution for the low frequency content and optimal time localization for the high frequency content was obtained. Slow events were analyzed with long windows, whereas faster events were analyzed with shorter windows. In addition, the continuous wavelet transform enables logarithmic frequency resolution. Programs for the calculation of wavelet transform and graphical presentations were written in MatLab (The MatWorks, MA). In the interval 0.0095–2 Hz, periodic oscillations with five different characteristic frequencies are observed (Fig. 1). The mean amplitude in this interval (denoted as mean amplitude) was first determined, and then the amplitude of each particular frequency interval (denoted as absolute amplitude) was calculated. The relative amplitude was then defined as the ratio between the absolute amplitude at a particular frequency interval and the mean amplitude of the entire spectrum (11). Spectral power analyses are not presented in this article.
The flow, wavelet, anthropometric, and hemodynamic data were all normally distributed and are presented as mean ± sd. The t test to compare dependent samples was used to evaluate differences in skin perfusion and oscillatory components obtained by wavelet transform in basal probes. Differences in the wavelet transform between ACh and SNP probes were evaluated by a paired t test. Data were analyzed with Sigmastat (Systat Software, Richmond, CA). P < 0.05 was considered statistically significant.
Effects of General Anesthesia on Hemodynamics, Respiration Frequency, and Skin Temperature
General anesthesia induced a reduction in systolic blood pressure by 16% (P = 0.002), in diastolic blood pressure by 15% (P = 0.01), and in mean arterial blood pressure by 16% (P = 0.007). There were no significant changes in heart rate. The respiratory rate applied by the ventilator was significantly slower than the spontaneous breathing rate before induction of general anesthesia (P = 0.025) (Table 2). After general anesthesia, skin temperature was not altered. In the control group, there were no significant differences in arterial blood pressure, respiration frequency, or in skin temperature between the two registrations.
Effects of General Anesthesia on Basal Skin Perfusion
During general anesthesia, basal skin perfusion, as measured by laser Doppler flowmetry, increased from 18 ± 8 to 25 ± 10 AU (P = 0.039) in probe 1 and from 18 ± 7 to 24 ± 8 AU (P = 0.026) in probe 2 (Fig. 2, panel A). In the control group, there were no significant differences in basal skin perfusion between the two registrations (probe 1: 18 ± 4 and 21 ± 8 AU, P = 0.15, and probe 2: 17 ± 6 and 16 ± 5 AU, P = 0.19).
Effects of General Anesthesia on Skin Perfusion in Response to Iontophoresis with ACh and SNP
Average skin perfusion in response to iontophoresis with ACh and SNP was not significantly altered by general anesthesia (probe 3-ACh: 65 ± 38 before and 65 ± 29 AU after anesthesia, P = 0.99; probe 4-SNP: 91 ± 34 before and 68 ± 23AU after anesthesia, P = 0.08) (Fig. 2, panel B). In the control group, there were no significant differences between the two registrations (ACh: 68 ± 21 and 63 ± 31 AU, P = 0.70; SNP: 70 ± 32 and 67 ± 24 AU, P = 0.81).
Effects of General Anesthesia on Spectral and Relative Amplitudes Within Each of the Five Frequency Intervals During Basal Perfusion
During general anesthesia, there were highly significant reductions of spectral amplitudes in the 0.0095–0.021, 0.021–0.052, and 0.052–0.15 Hz frequency intervals in both probes. There was an increase in the 0.15–0.6 Hz frequency interval in both probes (Fig. 3, panels A and B). The same pattern was also found for relative amplitudes (Fig. 3, panels C and D). In the control group, there were no significant differences in the spectral amplitude between the two registrations in either of the probes (Fig. 4, panels A and B). In the relative amplitudes, there was a 12% reduction (P = 0.03) in the 0.021–0.052 Hz frequency interval in probe 1 (Fig. 4, panel C). There were no differences in relative amplitude between the two registrations in probe 2 (Fig. 4, panel D).
Effects of General Anesthesia on Relative Amplitudes Within Each of the Five Frequency Intervals for the ACh/SNP Ratio
Relative amplitudes in the 0.0095–0.021 Hz frequency interval were significantly higher for ACh than for SNP, both before and during general anesthesia (P < 0.001 for both). There were no significant differences between ACh and SNP in the other frequency intervals. In the control group, relative amplitudes were significantly higher for ACh than for SNP in the 0.0095–0.02 Hz frequency interval in both registrations, whereas there were no differences in the other intervals.
The main finding in this study is that general anesthesia induces major changes in the oscillations of human skin microcirculation. The amplitude of the three slowest oscillatory components of the perfusion signal is clearly reduced, whereas the amplitude of the oscillation in the 0.15–0.6 Hz frequency interval is increased. A representative three-dimensional wavelet transform from one of the patients illustrates this finding (Fig. 5). These effects of general anesthesia on microvascular oscillations confirm previous results in animal studies, although the analyzing techniques were different (13,14).
The hemodynamic effects of general anesthesia were as expected. Arterial blood pressure was reduced after induction of anesthesia. Propofol, alone or in combination with fentanyl and midazolam, reduces arterial blood pressure, mainly due to a reduction of systemic nervous activity that decreases systemic vascular resistance (15). The heart rate was unchanged. This can mainly be explained by an inhibitory effect of propofol on the baroreflexes, reducing the ability to increase heart rate as a response to hypotension (16,17).
Skin blood perfusion increased approximately 30% after induction of anesthesia. This modest increase was probably related to low sympathetic tone in blood vessels in the skin of the forearm (6), compared with vessels in skeletal muscles or to skin areas containing arteriovenous anastomoses, such as hands and feet, where sympathetic tone is high. The increase in skin perfusion could contribute to the lack of increase in temperature found in our study, as we would expect an increase in skin temperature during general anesthesia due to redistribution of body heat from the core to peripheral tissues (18).
Wavelet analyses of the laser Doppler flowmetry signal revealed major changes during general anesthesia. The oscillations in the 0.6–2 Hz frequency interval are mediated by heartbeats and are usually small in a laser Doppler flowmetry signal. This was the only interval that was not significantly altered during general anesthesia. As both arterial blood pressure and pulse pressure decreased, a decrease in this oscillation was anticipated. However, general vasodilatations will make the vessels that are available for laser Doppler flowmetry measurement more responsive to cardiac pulsation, and thereby oppose the effect of the decreased pulse pressure. These effects might explain the unchanged oscillation in this frequency interval.
The oscillations in the 0.15–0.6 Hz frequency interval are created by respiratory changes in intrathoracic pressure that induce variations in cardiac filling, stroke volume, and systemic blood pressure. Since skin perfusion in the human forearm correlates to arterial blood pressure due to low sympathetic tone, these variations can be detected by laser Doppler flowmetry (6). Mechanical ventilation increases stroke volume variations more than during spontaneous breathing (19). This effect is demonstrated by increased amplitude in this frequency interval.
We found a significant reduction of the amplitude in the 0.052–0.15 Hz frequency interval. We relate this to an impairment of myogenic oscillations in the vascular wall (6). The existence of oscillations that originate from within the vascular smooth vessels is well documented (2). These rhythmical contractions are most likely caused by repetitive release and reuptake of Ca2+ from the sarcoplasmatic reticulum, which spreads and synchronizes between smooth muscle cells by gap junctions (20). The frequency interval of these myogenic oscillations has been investigated in several studies in human skin microcirculation, although indirectly, by excluding the effect of the sympathetic nervous system and respiration (5,6,21–24). Our results correlate with findings in animal studies proposing a direct inhibitory effect of propofol on vascular smooth muscle cells, which may contribute to the reduction of systemic vascular resistance (25–27). On the contrary, Robinson et al. (28) did not detect any significant vascular responses after propofol infusions into the human brachial artery.
The oscillations in the 0.021–0.052 Hz frequency interval were significantly reduced. We interpret this as a decrease in sympathetic nervous activity due to general anesthesia. These oscillations in skin microcirculation are created both by a direct local sympathetic effect, and indirectly by variations in arterial blood pressure induced by the sympathetic nervous system. The frequency and amplitude of these oscillations are influenced by temperature and emotional reactions. Thus, different experimental conditions could explain why sympathetic nervous activity has been related to a different frequency interval, often around 0.1 Hz, in other studies (29–31). Different spectral analyzing techniques and analysis of different parts of the cardiovascular system other than the skin microcirculation could also contribute to this finding. However, the connection between this oscillatory component and reduced sympathetic nerve activity in human skin microcirculation is supported by several studies. Kastrup et al. (5) found that oscillations, with amplitude around 0.025 Hz, disappeared completely after local and ganglion nerve block. Additionally, Söderström et al. (32) found that oscillations in the 0.021–0.052 Hz frequency interval were significantly reduced in laser Doppler flowmetry signals from human-free flaps. This was also confirmed in our recent brachial plexus block study. In this study, the oscillations related to myogenic activity were not altered (24). In the control group, there was a small but significant reduction in relative amplitude in the 0.021–0.052 Hz frequency interval in probe 1. However, this was not confirmed by the other probe.
In the 0.0095–0.052 Hz frequency interval there was a highly significant reduction in amplitude. We have previously demonstrated that the oscillations in this frequency interval could be related to the vascular endothelium, as iontophoresis with ACh increased relative amplitude more than SNP in this frequency interval (7). In a later study, we showed that this effect was inhibited by NG-monomethyl-l-arginine, an inhibitor of nitric oxide synthesis, and reversed by l-arginine, the substrate for nitric oxide synthesis (33). If the reduced amplitude of the oscillation in the 0.0095–0.052 Hz frequency interval represents an impairment of the vascular endothelium caused by general anesthesia, this should also be seen in the iontophoretic data. However, this was not the case. The difference seen in relative amplitude between ACh and SNP for the oscillations in the 0.0095–0.052 Hz frequency interval was not altered by general anesthesia (Fig. 3, panel B). There was also no significant difference in skin perfusion induced by ACh or SNP before and during general anesthesia (Fig. 2). An additional explanation could be that the frequency interval related to sympathetic activity is wider than was originally described by us and thus influences the 0.0095–0.052 Hz frequency interval.
LIMITATIONS OF THE STUDY
In this study, we used a combination of three anesthetic drugs (propofol, fentanyl, and midazolam), which is common in clinical practice. Even though propofol is the dominant drug, the findings in our study relate to the total effect of these drugs.
According to our hypothesis, we calculated sample size based on the anticipated reduction of sympathetic nervous activity. We found that with α = 0.05 and a desired power of 80%, we would need a minimum number of 10 patients to show a 50% decrease of the amplitude in the 0.021–0.052 Hz frequency interval. Because of many variables in our study, we cannot exclude the possibility of having type 2 errors in our statistical calculations, especially in variables with large variation, such as the iontophoretic data.
Variation in sympathetic tone throughout the day has the potential to influence our results. In the anesthesia group, all the recordings were obtained in the early morning. Thus, variations in sympathetic tone should have had little influence on data in this group. However, in the control group, recordings were performed both in the morning and the afternoon. This could cause greater variations in data, but should not have influenced the comparison between registration 1 and 2.
Alterations in skin microcirculation induced by general anesthesia can be evaluated by wavelet transform of the laser Doppler signal. General anesthesia reduced the three slowest oscillatory components of the perfusion signal related to inhibition of sympathetic and myogenic activity and endothelial function. However, the iontophoretic data did not reveal a selective effect on the endothelium. The increase in the 0.15–0.6 Hz frequency interval relates to the effect of mechanical ventilation on stroke volume variation. In our opinion, continuous wavelet transform with logarithmic frequency resolution of laser Doppler flowmetry signals could be a valuable research tool for studying microcirculatory changes in clinical experiments.
This project was funded by the Norwegian Anesthesiological Association, Surgical Division of Ulleval University Hospital, Oslo, Norway, Slovenian Research Agency, and the FP6 EU NEST-Pathfinder project: “BRACCIA” (Brain, respiratory and cardiac causalities in anesthesia).
1.Jones TW. Discovery that the veins of the bat's wing are endowed with rhythmical contractility and that onward flow of blood is accelerated by each contraction. Philos Trans R Soc Lond 2006;142:131–6
2.Aalkaer C. Vasomotion: cellular background for the oscillator and for the synchronization of smooth muscle cells. Br J Pharmacol 2005;144:605–16
3.Tsai AG. Evidence of flowmotion induced changes in local tissue oxygenation. Int J Microcirc Clin Exp 1993;12:75–88
4.Stefanovska A, Bracic M, Kvernmo HD. Wavelet analysis of oscillations in the peripheral blood circulation measured by laser Doppler technique. IEEE Trans Biomed Engl 1999;46: 1230–9
5.Kastrup J, Bulow J, Lassen NA. Vasomotion in human skin before and after local heating recorded with laser Doppler flowmetry. A method for induction of vasomotion. Int J Microcirc Clin Exp 1989;8:205–15
6.Lossius K, Eriksen M. Spontaneous flow waves detected by laser Doppler in human skin. Microvasc Res 1995;50:94–104
7.Kvernmo HD, Stefanovska A, Kirkeboen KA, Kvernebo K. Oscillations in the human cutaneous blood perfusion signal modified by endothelium-dependent and endothelium-independent vasodilators. Microvasc Res 1999;57:298–309
8.Nilsson GE, Tenland T, Oberg PA. Evaluation of a laser Doppler flowmeter for measurement of tissue blood flow. IEEE Trans Biomed Engl 1980;27:597–604
9.Bollinger A, Hoffmann U, Franzeck UK. Evaluation of flux motion in man by the laser Doppler technique. Blood Vessels 1991;28:21–6
10.Westerman RA, Widdop RE, Hogan C, Zimmet P. Non-invasive tests of neurovascular function: reduced responses in diabetes mellitus. Neurosci Lett 1987;81:177–82
11.Bracic M, Stefanovska A. Wavelet-based analysis of human blood-flow dynamics. Bull Math Biol 1998;60:919–35
12.Grossmann A, Morlet J. Decomposition of Hardy functions into square integrable wavelets of constant shape. SIAM J Math Anal 1984;4:723–36
13.Colantuoni A, Bertuglia S, Intaglietta M. Effects of anesthesia on the spontaneous activity of the microvasculature. Int J Microcirc Clin Exp 1984;3:13–28
14.Hundley WG, Renaldo GJ, Levasseur JE, Kontos HA. Vasomotion in cerebral microcirculation of awake rabbits. Am J Physiol 1988;254:H67–71
15.Grounds RM, Twigley AJ, Carli F, Whitwam JG, Morgan M. The haemodynamic effects of intravenous induction. Comparison of the effects of thiopentone and propofol. Anaesthesia 1985;40: 735–40
16.Ebert TJ. Sympathetic and hemodynamic effects of moderate and deep sedation with propofol in humans. Anesthesiology 2005;103:20–4
17.Sellgren J, Ejnell H, Elam M, Ponten J, Walin BG. Sympathetic muscle nerve activity, peripheral blood flows, and baroreceptor reflexes in humans during propofol anesthesia and surgery. Anesthesiology 1994;80:534–44
18.Matsukawa T, Sessler DI, Sessler AM, Schroeder M, Ozaki M, Kurz A, Cheng C. Heat flow and distribution during induction of general anesthesia. Anesthesiology 1995;82:662–73
19.Michard F. Changes in arterial pressure during mechanical ventilation. Anesthesiology 2005;103:419–28
20.Peng H. Hypothesis for the initiation of vasomotion. Circ Res 2001;88:810–15
21.Salerud EG. Rhythmical variations in human skin blood flow. Int J Microcirc Clin Exp 1983;2:91–102
22.Smits TM. Hyperventilation-induced changes in periodic oscillations in forehead skin blood flow measured by laser Doppler flowmetry. Int J Microcirc Clin Exp 1987;6:149–59
23.Wilkin JK. Periodic cutaneous blood flow during postocclusive reactive hyperemia. Am J Physiol 1986;250:8
24.Landsverk SA, Kvandal P, Kjelstrup T, Benko U, Bernjak A, Stefanovska A, Kvernmo H, Kirkeboen KA. Human skin microcirculation after brachial plexus block evaluated by wavelet transform of the laser Doppler flowmetry signal. Anesthesiology 2006;105:478–84
25.Imura N, Shiraishi Y, Katsuya H, Itoh T. Effect of propofol on norepinephrine-induced increases in [Ca2+
]i and force in smooth muscle of the rabbit mesenteric resistance artery. Anesthesiology 1998;88:1566–78
26.MacPherson RD, Rasiah RL, McLeod LJ. Propofol attenuates the myogenic response of vascular smooth muscle. Anesth Analg 1993;76:822–9
27.Chang KS. Propofol produces endothelium-independent vasodilation and may act as a Ca2+
channel blocker. Anesth Analg 1993;76:24–32
28.Robinson BJ, Ebert TJ, O'Brien TJ, Colinco MD, Muzi M. Mechanisms whereby propofol mediates peripheral vasodilation in humans. Sympathoinhibition or direct vascular relaxation? Anesthesiology 1997;86:64–72
29.Malpas SC. Neural influences on cardiovascular variability: possibilities and pitfalls. Am J Physiol Heart Circ Physiol 2002;282:H6–20
30.Malliani A. Cardiovascular neural regulation explored in the frequency domain. Circulation 1991;84:482–92
31.Bernardi L. Synchronous and baroceptor-sensitive oscillations in skin microcirculation: evidence for central autonomic control. Am J Physiol 1997;273:78
32.Söderström T, Stefanovska A, Veber M, Svensson H. Involvement of sympathetic nerve activity in skin blood flow oscillations in humans. Am J Physiol Heart Circ Physiol 2003;284:H1638–46
33.Kvandal P, Stefanovska A, Veber M, Kvernmo H, Kirkeboen KA. Regulation of human cutaneous circulation evaluated by laser Doppler flowmetry, iontophoresis, and spectral analysis: importance of nitric oxide and prostaglandines. Microvasc Res 2003;65:160–71