|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1Department of Internal Medicine, University of Schleswig Holstein, Luebeck, Germany; 2Department of Internal Medicine, Lundberg Laboratory for Diabetes Research; and 3Department of Clinical Neurophysiology, University of Goteborg, Gothenberg, Sweden
Submitted 14 March 2006 ; accepted in final form 14 June 2006
| ABSTRACT |
|---|
|
|
|---|
blood flow; human; autonomic nervous system
In humans, ATBF in vivo has preferentially been measured using tracer washout techniques (26) or microdialysis methods based on ethanol dilution (13, 14, 17). Although some authors find a good correlation of both methods, others report substantial differences between both techniques, e.g., during the measurement of adipose tissue perfusion (21). Additionally, both methods are limited by a low temporal resolution (30). An alternative method is needed for the assessment of rapid, neurally evoked local perfusion changes.
Laser-Doppler flowmetry (LDF) was first described in 1975 by Stern (39) and has been proven useful for perfusion measurements in virtually every organ system (20, 28, 33, 40, 41). Compared with the tracer washout or microdialysis methods mentioned above, the LDF technique is easily performed and minimally invasive and has an excellent time resolution. In brief, a beam of laser light, carried by a fiber-optic probe attached to or inserted into the investigated tissue, hits moving blood cells and changes its wavelength (Doppler shift). The magnitude and frequency distribution of these changes in wavelength are directly related to the number and velocity of blood cells. The information is picked up by a returning fiber, converted into an electronic signal, and analyzed. Only a few studies have evaluated the feasibility of LDF in adipose tissue (23, 25).
Despite their close anatomical relation, skin and subcutaneous adipose tissue must be regarded as two distinct organs, differing in thermoregulatory and metabolic capacity as well as in neural/endocrine regulation. Skin blood vessels and sweat glands are generally considered to be under pure sympathetic (both adrenergic and cholinergic) control (11, 19), but a parasympathetic innervation of cutaneous blood vessels has also been proposed (36). Adipose tissue innervation is sparse and generally considered to be of sympathetic origin (10, 12, 34), but recent experiments in animals suggest additional parasympathetic innervation of adipose tissue (4). Hence, physiological maneuvers affecting the autonomic nervous system may affect perfusion in skin and subcutaneous adipose tissue differently.
The primary aim of the present study was to assess LDF for monitoring ATBF and to discriminate between changes of ATBF and skin blood flow (SBF). For this purpose, rapid LDF-ATBF responses to standardized stimuli known to affect autonomic nervous function were assessed in lean females and compared with LDF-SBF responses obtained in parallel in an overlying skin area. Subsequently, the same maneuvers were performed in a group of obese but otherwise healthy female subjects to evaluate whether their ATBF regulation differs from that in subjects with normal body weight. Finally, LDF-SBF and -ATBF were compared with perfusion measured by the 133xenon tracer washout method during prolonged cooling in lean healthy subjects.
| METHODS |
|---|
|
|
|---|
The local ethics committees in Luebeck and Gothenberg approved the study, and all participants gave written informed consent before the experiments.
LDF and 133Xe Washout During Prolonged Cooling
In the first experiment, 10 lean healthy subjects [4/6 male/female; age range 2326 yr; body mass index (BMI) 22, 9 ± 0, 8 kg/m2] were investigated in the supine position at a room temperature of 25 ± 1°C and dressed in a thermal suit, used for subsequent cooling of the subjects. Two noninvasive LDF probes were attached to the skin on the thigh and the belly below the umbilicus, respectively, for recording skin perfusion. In the same two regions, two invasive LDF probes were inserted into the subcutaneous adipose tissue via plastic catheters for recording adipose tissue perfusion. Depots of 133Xe were injected subcutaneously in two adjacent regions for monitoring 133Xe washout. After a 45-min equilibration period following placement of LDF probes and tracer depots, LDF and tracer dilution was monitored for 30 min in a warm (25 ± 1°C) condition and for 60 min during cooling, induced by circulating cold (1218°C) water through the thermal suit. Relative changes in 133Xe washout during the first and second one-half hour of cooling were compared with relative changes in skin and adipose tissue LDF perfusion, respectively. Periods with artifacts due to cooling-induced shivering as well as LDF records exhibiting signal shifts, indicating altered probe position, were excluded from analysis (see below).
ATBF and SBF During Short Maneuvers in Lean and Obese
Eleven lean and eleven obese age-matched (2239 yr) female volunteers were recruited. Mean BMI was 21.7 ± 0.52 kg/m2 (range 2023 kg/m2) in the lean and 33.4 ± 1.23 kg/m2 (range 3135 kg/m2) in the obese subjects, respectively. Their body weight had remained stable for 3 mo before the examination.
Heart rate was recorded online using a three-lead ECG, and breathing was monitored by a respiration transducer (Pneumotrace 1130; all ADInstruments, Spechbach, Germany). Finger blood pressure (fBP) was measured continuously using the volume clamp method throughout the experiment (Finapres, Ohmeda 2600, Engelwood, CO). In addition, oscillometric blood pressure (BP) was measured automatically (Welch Allyn, Skaneateles Falls, NY) from the opposite arm. Skin temperature was continuously measured in the vicinity of the LDF probe and stored on a mini-logger (series 2000; Mini-Mitter, Bend, OR).
LDF was performed with a two-channel Periflux system (Perimed, Jarfalla, Sweden); one channel was used for subcutaneous measurement and the other for assessment of skin blood flow. After a warm-up phase of 30 min, laser-Doppler flow probes were calibrated with motility standard. Subsequently, a needle probe (Perimed) was inserted through a plastic cannula (Becton Dickinson, Heidelberg, Germany) into the lateral femoral subcutaneous adipose tissue. To minimize movement artifacts, the position of the cannula was secured with tape. The skin probe was attached to the skin 10 mm apart from the tip of the subcutaneous needle probe.
ECG, continuous blood pressure curves, respiration, and LDF signals were registered online with a multichannel analog-digital converter (PowerLab sp16, ADInstruments) at a sampling rate of 200/s.
Protocol
All experiments were performed in a quiet room, protected from outside noise. Subjects arrived at the research unit after an overnight fast. They rested comfortably in a supine position, with the upper body slightly (30°) elevated. The lower half of the body was placed in a lower body negative pressure (LBNP) box (see below). The subcutaneous LDF probe was inserted, and the cutaneous LDF probe was attached. ECG electrodes, respiration transducer, fBP cuff, oscillometric BP cuff, and the temperature probe were applied.
All volunteers were exposed to the different maneuvers in a fixed sequence, starting with an auditory startle stimulus and followed by a Valsalva-like straining maneuver and a LBNP maneuver. Maneuvers were separated by intervals of at least 5 min and were preceded by a 2-min baseline period.
Auditory startle. The startle stimulus, presented without prior warning, consisted of an unexpected loud and short noise. This maneuver caused sudden involuntary movements in most subjects, associated with artificial flow curves and subsequent baseline changes in LDF signals derived from the adipose tissue. Hence, the auditory startle maneuver was not analyzed further.
Straining maneuver. Like the Valsalva, this maneuver causes a complex pattern of changes in heart rate and BP that can be divided in four distinct consecutive phases representing both mechanical and reflex mechanisms (5). Our subjects were asked to strain against a closed glottis with maximal strength for 20 s. A second identical straining maneuver was performed after a resting period of 5 min. The maneuver with the highest increase in heart rate during phase II was selected for further analysis. Laser-Doppler signal analysis included a period beginning 10 s before straining until 90 s after.
LBNP. The LBNP box encased the lower part of the body up to the iliac crest. The box was closed, and the opening for the abdomen was sealed with a flexible, broad plastic skirt to allow the application of negative pressure. A standard household vacuum cleaner was used to apply suction of 15 mmHg, which was determined by a pressure transducer. An adjustable footboard prevented subjects from sliding into the box during suction. Suction was applied for 10 min, and signal analysis began 100 s before suction and continued until 4 min after the end of LBNP.
Statistical Evaluation
LDF probes are sensitive to movement, and periods with artifacts in flow curves have to be excluded from analysis. Artifacts are characterized by sudden sharp up- or downward deflections of the perfusion curve or a chaotic, pulse wave-independent signal. Such periods were identified visually and excluded from further analysis. When the flow signal after such sudden curve deviations did not return to the baseline level, an alteration of the position of the flow probe was assumed, and the data of the complete maneuver were discarded. For intraindividual comparisons, baseline values 60 s before the respective maneuver were averaged and set at 100%. The following values were expressed as percent baseline.
Statistical analysis of the effects of short maneuvers relied on ANOVA with a repeated-measures factor "time," factor "location" (subcutaneous vs. cutaneous blood flow), and factor "body weight" (lean vs. obese). When ANOVA identified significant effects of the factors, post hoc pairwise comparisons were performed subsequently. Degrees of freedom were corrected according to the method of Greenhouse-Geisser. Results are presented as means ± SE. Changes in LDF-ATBF, LDF-SBF, and 133Xe washout during prolonged cooling were evaluated with linear regression analysis. P < 0.05 was considered significant.
| RESULTS |
|---|
|
|
|---|
Although adequate LDF perfusion signals were achieved in all subjects while they were resting at a comfortable, warm room temperature, cooling induced periods of shivering in many subjects, and this occasionally led to altered probe positions. Hence, the comparison of relative changes in 133Xe washout and LDF perfusion is based on eight recordings from the thigh and six from the belly (4 subjects contributing data from both thigh and belly, 6 subjects contributing data from thigh or belly only). Cooling induced a reduction in 133Xe washout (i.e., in ATBF) from 10 of 14 recording sites, a reduction of LDF perfusion in all skin sites, and a reduction of LDF perfusion in 12 of 14 subcutaneous probe sites. Relative changes in 133Xe washout showed a positive correlation with changes in skin LDF (r = 0.694, P < 0.001; Fig. 1A) and subcutaneous LDF (r = 0.506, P < 0.01; Fig. 1B).
|
Straining maneuver. Whereas the four distinct phases of hemodynamic changes elicited by straining were discernible also in SBF and ATBF (Fig. 2), the perfusion changes differed markedly between SBF and ATBF.
|
As expected, heart rate rose significantly during phase II (Fig. 2D) and BP fell (Fig. 2C), with this decline being more pronounced in lean subjects (P < 0.05 for factor "weight"). ATBF as well as SBF fell in parallel to the decrease in mean BP in both weight groups. In the obese group, however, ATBF remained on a higher level compared with SBF (P < 0.05 for the factor location).
During phase III, immediately after the termination of straining, SBF and ATBF started to rise again both in lean and obese subjects (P < 0.001 for the factor time).
During phase IV, SBF exceeded baseline perfusion and returned to baseline within 20 s in both groups. In contrast, ATBF returned to baseline immediately after the end of phase III and did not increase further in lean subjects. Obese subjects, on the other hand, exhibited a prolonged increase in ATBF, resulting in a higher adipose tissue perfusion in obese vs. lean subjects after 80 s (Fig. 2A; P < 0.05).
LBNP. LBNP of 15 mmHg for 10 min caused an immediate and similar decrease in SBF and ATBF in both lean and obese subjects (P < 0.001 for the factor time) that remained constant throughout the period of negative pressure application (Fig. 3). After suction was terminated, SBF rapidly returned to baseline values. Whereas ATBF in lean volunteers paralleled SBF throughout the whole maneuver, ATBF in obese subjects rose above baseline after termination of LBNP and was significantly higher than ATBF in the lean group (P < 0.05 for the factor weight). LBNP had no effect on mean BP or heart rate in lean and obese groups.
|
| DISCUSSION |
|---|
|
|
|---|
Laser Doppler has been used to evaluate local perfusion in a variety of tissues such as the brain, gut, bone, skin, and muscle (13, 9, 15, 16, 18, 24, 27). Many authors have found that perfusion changes estimated with LDF compare with those obtained through 133Xe washout (22, 31, 32), and this was confirmed in the present study, where both subcutaneous and skin LDF correlated with 133Xe washout from a subcutaneous depot during prolonged whole body cooling. In fact, 133Xe washout was more closely related to cutaneous than to subcutaneous LDF perfusion. This may indicate that a subcutaneous xenon depot is not solely drained through the local adipose tissue microcirculation but also through the capillary network of the epidermal layer. The epidermis is highly perfused and may even be more important for the drainage of the superficial portion of the xenon depot. If so, perfusion measurements with a subcutaneously inserted LDF probe offer the advantage of restricting flow measurements to the fat tissue compartment. This feature may be particularly important when changes in ATBF are not concordant with SBF changes, as observed during short maneuvers in the present study. An additional advantage of the LDF method is its high time resolution: the rapid sequence of perfusion changes during the straining maneuver, with a sudden increase (phase I) and a subsequent decrease (phases II and III) within 30 s, is unlikely to be detected at all with the 133Xe washout technique, not only because the duration of changes is too short but also because the net effect of opposite flow changes measured with a dilution method may not yield any net change in perfusion at all.
The advantages of subcutaneous LDF, such as high time resolution and tissue-specific measurement, have to be weighed against two major drawbacks. First, subcutaneous LDF probes were particularly sensitive to movement artifacts. In the present study, the LDF probe was fixed to the skin, and subjects were carefully instructed to avoid movements during the experiment. In all flow curves, every abrupt change was considered to be a movement artifact and excluded from analysis. According to this strict approach, we had to abandon the analysis of flow responses to the acoustic startle maneuver in all subjects, since this stimulus consistently induced movement artifacts and subsequent baseline shifts. The second disadvantage of the method is that it does not allow any absolute measurement of blood flow, since the method does not measure linear flow a priori. Therefore, we only report relative perfusion changes in relation to a stable baseline period before the maneuver.
Another important methodological aspect is the need for continuous blood pressure measurement during LDF recordings of local perfusion, given the complex relationships between arterial blood pressure, peripheral resistance, and peripheral tissue perfusion. In the present study we tried to dissociate influences of blood pressure changes and autonomic inputs to the adipose tissue and skin microcirculation during mild LBNP, a well-established model for sympathetic activation via the unloading of cardiopulmonary baroreceptors, without affecting arterial blood pressure (7, 8). LBNP induced a prompt decrease in blood flow, which was equivalent in femoral skin and adipose tissue, without associated changes in blood pressure or heart rate. The perfusion changes did not differ between lean and obese subjects during suction. This suggests that the microcirculation of both compartments is under active sympathetic control during this maneuver, in both lean and obese subjects (29, 35). The reduction in femoral ATBF in the present study was comparable with the 34% decrease in forearm ATBF during LBNP previously demonstrated with 133Xe washout (38).
At first glance, body weight had only minor effects on stimulus-induced ATBF, indicating that the regulation of blood flow in subcutaneous fat tissue is not dramatically changed in obesity. However, after LBNP or straining, a prolonged ATBF increase was only observed in obese subjects. In contrast, vasoconstrictor responses during these maneuvers were not enhanced in obese subjects. Thus our study suggests that active vasodilatation or, alternatively, passive vasodilatation after a vasoconstriction is increased in the subcutaneous adipose tissue of obese subjects. Notably, a decreased ATBF and a blunted response to oral glucose have been demonstrated in obese humans (9, 21). Taken together with the data presented, this may indicate that maximal vasodilatation is achieved under basal conditions in the obese and regained after vasoconstriction. Further studies combining LDF-ATBF with subcutaneous microdialysis (21), i.e., during metabolic stimulation, are warranted to explore adipose tissue metabolism in the obese.
Resembling the Valsalva maneuver, straining elicits mechanical and reflex effects that involve both sympathetic and parasympathetic neuronal activity (5, 6, 37). In the present study, this complex maneuver revealed that skin and adipose tissue perfusion shows partly parallel but partly opposite changes within a short time frame, in all probability due to tissue-specific differences in the impact of sympathetic (and possibly also parasympathetic) reflex mechanisms. However, the exact autonomic regulatory mechanism mediating the observed perfusion changes during straining cannot be characterized on the basis of the present study. For this purpose, LDF needs to be monitored while appropriate pharmacological agents selectively block different autonomic efferent effects.
In conclusion, the present study demonstrates that LDF is a useful tool to determine specifically blood flow changes in subcutaneous adipose tissue of nonmoving subjects. In situations of static changes of tissue perfusion, it compares well with other established methods like 133Xe washout. However, in addition to its local specificity, it has the advantage of detecting rapid perfusion changes caused by dynamic autonomic control. Disadvantages are the propensity for movement artifacts and the inability to yield absolute blood flow values. On the basis of parallel measurements in skin and adipose tissue in the same innervation area, we were able to show that blood flow in these compartments can be differentially regulated. Finally, our study demonstrates that vasodilatation after phases of vasoconstriction seems to be increased in obese subjects.
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
| REFERENCES |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |