| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1Institute for Exercise and Environmental Medicine, Presbyterian Hospital of Dallas, and 2Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, Texas
Submitted 10 January 2005 ; accepted in final form 15 November 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
sympathetic nervous system; autonomic; thermoregulation; variability
SSNA continuously exhibits rhythmic changes. When SSNA was first reported by Hagbarth et al. (7), a loose coupling between spontaneous SSNA and resting respiratory rhythm was observed in some subjects. Respiratory rhythms in integrated SSNA (1) and in single vasoconstrictor and sudomotor neurons (13) were verified using interval histogram analytic techniques. Using spectral analysis of integrated SSNA, Cogliati et al. (2) showed that SSNA during normothermia had a high-frequency (HF) oscillatory component, which reflected the respiratory rhythm. Moreover, they reported that SSNA also had a low-frequency (LF, 0.1-Hz) oscillatory component, which was coherent with LF components of R-R interval and blood pressure (2).
Exposure of humans to hyperthermia leads to pronounced increases in skin blood flow, sweat rate, and SSNA (1, 4, 26). These responses likely originate from increases in sudomotor and vasodilator activities (1, 18), which are different from vasoconstrictor activity during normothermia and cooling. Bini et al. (1) reported that when sweating was more pronounced in a hot environment, SSNA burst duration was narrower than that observed in normothermic conditions and that the interval between SSNA bursts became shorter. However, the effects of heat stress on frequency components of SSNA are unknown. Such information may provide insight into the control mechanisms of SSNA in heat-stressed subjects. Because the main components of SSNA under heat stress are different from those during normothermia and cooling (1, 18), we hypothesize that spectral characteristics of SSNA in heat-stressed subjects will be different from those in subjects subjected to normothermia and cooling. To test this hypothesis, the spectrum of integrated SSNA in humans was evaluated during normothermia, cooling, and heat stress.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Seventeen healthy subjects (10 men and 7 women) participated in the study. The subjects' average age was 32 ± 2 (SE) yr, and all were of normal height (172 ± 2 cm) and weight (72 ± 3 kg). Subjects were normotensive (supine blood pressures <140/90 mmHg), were not taking medications, and were free of any known cardiovascular, neurological, and metabolic diseases. All subjects abstained from caffeine, alcohol, and exercise 24 h before the study. A written informed consent, which was approved by the Institutional Review Boards at the University of Texas Southwestern Medical Center and Presbyterian Hospital of Dallas, was obtained from each subject before participation in the study.
Instrumentation and Measurements
Internal (i.e., core) temperature (Tco) was indexed from an ingestible pill telemetry system (HTI Technologies, Palmetto, FL; n = 7) or from a thermocouple placed in the sublingual sulcus (n = 10). Telemetry pill measurements correlate well with other internal temperature measurements such as esophageal temperature (14). Mean skin temperature (
sk) was measured via the weighted average of six thermocouples attached to the skin (20). Mean body temperature was calculated as follows: 0.9·Tco + 0.1·sk. Each subject was dressed in a tube-lined suit that permitted control of sk by change in the temperature of the water perfusing the suit.
Multifiber recordings of SSNA were obtained with a tungsten microelectrode inserted into the common peroneal nerve. A reference electrode was placed subcutaneously 2–3 cm from the recording electrode. The recording electrode was adjusted until a site was found in which SSNA bursts were clearly identified using previously established criteria (6, 22): 1) integrated nerve activity nonsynchronous with the heartbeat, 2) irregular burst activity, 3) generation of reflex bursts during mental or somatosensory stimuli (e.g., loud sound and light stroking of the innervated region), and 4) absence of an increase in activity during inspiratory apnea. The nerve signal was amplified, passed through a band-pass filter with a bandwidth of 500–5,000 Hz, and integrated with a time constant of 0.1 s (Iowa Bioengineering, Iowa City, IA). Mean voltage neurograms were displayed on a chart recorder. The nerve signal was also routed to an oscilloscope and a loudspeaker for monitoring throughout the study.
Blood pressure was recorded on a beat-by-beat basis from a finger via a Finapres device (Ohmeda, Louisville, CO). Resting blood pressures obtained from the Finapres device were verified during the experiment by auscultation of the brachial artery (SunTech Medical Instruments, Raleigh, NC). Respiratory excursions were monitored with piezoelectric pneumography (model 1132 Pneumotrace II). Skin blood flow was measured via laser-Doppler flowmetry using an integrating flow probe with a time constant of 0.1 s (Perimed, North Royalton, OH) from the area within the innervation field of the SSNA being recorded (usually the dorsal foot). The integrating flow probe consists of seven pairs of optical fibers and measures skin blood flow continuously from a 
28-mm2 area. After the entire procedure, a 3-cm-diameter heater element (Perimed), which housed the laser-Doppler flow probe, was engaged to elevate local sk to 42°C. Local temperature was held at this level for 30 min to elicit maximal cutaneous vasodilation (21). Sweat rate was also measured within the innervation field of the SSNA being recorded via capacitance hygrometry (Viasala, Woburn, MA) by perfusion of 100% nitrogen at a flow rate of 300 ml/min through a ventilated capsule (2.83-cm2 surface area) attached to the surface of the skin.
Protocols
Protocol 1: whole body heating.
All variables were recorded for 6 min in 10 subjects resting in the supine position with sk clamped by perfusion of normothermic (i.e., 34°C) water through the tube-lined suit. After normothermic data collection, sk was gradually increased to 38°C by perfusion of the tube-lined suit with 46°C water. Once Tco increased 0.5–0.7°C, the temperature of the water was reduced to 44–45°C to attenuate the rate of rise of internal temperature throughout the ensuing data collection period. In this heat-stressed condition, data were collected for an additional 6 min. Respiratory frequency was not controlled in either thermal condition.
Protocol 2: whole body cooling and whole body heating.
Whole body cooling increases cutaneous vasoconstrictor activity. To identify the spectral characteristics of SSNA during skin cooling, SSNA was obtained and analyzed from an additional seven subjects during normothermia and cooling. All variables were recorded for 6 min from the normothermic subjects as described in protocol 1. After normothermic data collection, 15°C water was perfused through the suit. After 7–10 min, a 2–3°C decrease in sk was accompanied by clear increases in SSNA. Under this condition, all variables were recorded for another 6 min. sk was then returned to baseline. The total period of cooling (sk < baseline) was <20 min. After 5 min, the normothermic subjects were subjected to whole body heating as described in protocol 1. After Tco increased 0.5–0.7°C, data were collected from these subjects for an additional 6 min. Because the cooling was moderate and short and the normothermic and heat stress conditions in protocol 2 were controlled in a manner similar to that described for protocol 1, data for the normothermic and heat-stressed conditions from these 7 subjects were combined with normothermic and heat-stress data from the 10 subjects in protocol 1, resulting in data from 17 subjects being analyzed during normothermia and heat stress.
Data Analysis
Data were sampled at 200 Hz via a commercial data acquisition system (Biopac System, Santa Barbara, CA) and analyzed using LabView software (National Instruments, Austin, TX). SSNA bursts were first identified in real time by visual inspection of data plotted on a chart recorder coupled with the burst sound from the audio amplifier. SSNA recordings with indications of electrode movement were excluded from the analysis. Integrated SSNA was normalized by assignment of a value of 100 to the mean amplitude of the largest sympathetic bursts (top 10% of identified bursts) during normothermic baseline (5, 8, 19), and subsequent bursts in the neurogram were normalized against that value. This form of normalization was employed to reduce variability in the SSNA recordings attributed primarily to recording electrode positioning between subjects. To assess total activity of SSNA, the baseline was carefully identified, and the area of the integrated neurogram above this baseline was measured from the digitized record. Beat-by-beat mean arterial blood pressure (MAP), R-R interval, and skin blood flow were calculated. The respiratory trace was normalized by assignment of a value of 100 to the mean amplitude during the 6-min normothermic baseline period.
After calculating the autopower of SSNA using different sampling frequencies, we found that most of the SSNA spectral power was in the <2.5-Hz frequency band. To estimate the autopower of SSNA, the original 200-Hz neurogram recordings were resampled at 8 Hz from the mean value of each set of 25 data points. Respiration traces were resampled using the same method. Beat-by-beat data series of R-R interval, MAP, and skin blood flow were interpolated (cubic spline) and resampled at 8 Hz. The autopower spectra of the resampled data were calculated via the Welch method (25) using Matlab software (Math Works, Natick, MA). In all thermal conditions, data were subdivided into 1,024-point segments (128 s) with 50% overlap, windowed (Hanning method), transformed, and averaged. For consistency with previous studies (2, 5), the LF (0.03- to 0.15-Hz) and HF (0.15- to 0.45-Hz) spectral powers of SSNA were calculated from the autospectra. Because the present results showed that spectral power of integrated SSNA can be >0.45 Hz but <2.5 Hz, we calculated the spectral power between 0.45 and 2.5 Hz and termed this the very-high-frequency (VHF) component. We calculated the LF and HF spectral power of R-R interval, MAP, and skin blood flow from the autospectra. To be consistent with SSNA, we also calculated spectral powers in the VHF regions for R-R interval, beat-by-beat MAP, and skin blood flow, although spectral power of these variables in the VHF region was very low. Analysis of the very-low-frequency (0.00- to 0.03-Hz) component requires specific algorithms and longer data series and, thus, was not addressed in the present study.
The squared coherence function (K2) between SSNA and R-R interval, MAP, respiratory activity, or skin blood flow was calculated as the square of the cross-spectral density divided by the product of the individual power spectral densities. Coherence is a measure of the statistical link between two variability series at any given frequency and is expressed as a number between 0 and 1. As observed by Cogliati et al. (2), SSNA and systolic blood pressure, R-R interval, or respiratory activity showed high coherence only in a narrow frequency range, whereas respiratory rate varied among subjects in the present study. Therefore, we used the coherence value near the central frequency of the LF and HF segments as an indicator of the link between SSNA and the hemodynamic variability.
Paired t-tests were used to assess differences between normothermia and heat stress or between normothermia and cooling for most variables. Differences in the change of SSNA spectral power between LF, HF, and VHF bands with whole body heating or cooling were evaluated using a repeated-measures one-way ANOVA. P < 0.05 was considered significant. Values are means ± SE.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Whole body heating increased sk 3.2°C and Tco 0.7°C, resulting in significant increases in heart rate, skin blood flow, and sweat rate (Table 1). These data verify that the subjects were in a heat-stressed condition. In normothermia, SSNA varied among subjects. SSNA was very low in some subjects and relatively high in others (Fig. 1). In all cases, SSNA increased significantly with whole body heating (Table 1, Fig. 1).
|
|
|
|
|
Whole body cooling significantly decreased sk but did not significantly alter Tco (Table 1). No shivering was observed. Total SSNA increased significantly with cooling (Table 1). A significant decrease in skin blood flow (Table 1), coupled with increases in SSNA, indicates that cooling induced cutaneous vasoconstriction. Cooling increased the total power of SSNA variability in all regions (Fig. 5A), but, in contrast to heat stress, the relative increase of the power in the VHF region was not greater than the increase of the power in the LF and HF regions (cf. Figs. 3B and 5B). Moreover, the power distribution during cooling was not significantly different from that during normothermia (Fig. 4B).
|
sk had increased to 36°C but Tco had not yet changed, total SSNA decreased to a level lower than precooling baseline (206 ± 56 vs. 394 ± 91 units/min, P < 0.05). As heating continued and Tco began to rise, total SSNA progressively increased.
|
Whole body heating significantly increased heart rate but did not change mean blood pressure; however, respiratory rate was slightly, but significantly, elevated (Table 1). Whole body cooling significantly increased blood pressure but did not change heart rate or respiratory rate (Table 1).
For R-R interval variability, the spectral power of LF and HF oscillatory components, as well as total power, was significantly reduced by whole body heating, whereas the LF-to-HF ratio was elevated (Table 2). Moreover, the spectral power of the LF and HF oscillatory components of MAP was significantly attenuated by heat stress (Table 2). The peak frequency of respiration increased slightly but significantly after whole body heating. The variability of skin blood flow on the dorsal foot was increased significantly by whole body heating (Table 2). Cooling did not significantly alter the spectral power of the R-R interval or beat-by-beat MAP variability (Table 2). In the VHF region, spectral powers of R-R interval, beat-by-beat MAP, and skin blood flow were very low relative to the total power during normothermia, cooling, and heat stress (Table 2).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
The present data show that the spectral power of integrated SSNA distributes from 0 to 2.5 Hz during normothermia, cooling, and heat stress. This observation, from frequency domain analysis, is consistent with SSNA observations analyzed in the time domain (1). Using interval histograms, Bini et al. (1) showed that the interval between SSNA bursts varied from 0.4 to 4–5 s, being somewhat dependent on the thermal condition. In a subsequent study, Cogliati et al. (2) showed LF and HF oscillations of SSNA during normothermia, resulting in clear peaks in the SSNA spectral density curve at each of these regions. In the present study, although SSNA spectral power distributes across LF and HF regions in normothermia, SSNA spectral density did not exhibit a clear peak in the LF region (Fig. 2). An explanation for differences between the present findings and those of Cogliati et al. is not readily apparent. However, these differences may be related to the analytic methods used in the respective studies. The SSNA spectrum in the study of Cogliati et al. was evaluated with autoregressive parametric spectral algorithms, which allow the shape of the spectral density curve to be influenced by the order of the model. The Welch method used in present study is based on fast Fourier transform algorithms (nonparametric method) (25), in which the shape of the spectral density curve is not influenced by the parameters of the algorithms.
To increase skin blood flow and sweat rate, SSNA is greatly activated under hyperthermic conditions (1, 4, 26). In the present study, whole body heating significantly increased sk and Tco, skin blood flow, sweat rate, and SSNA. Spectral analysis revealed that the variability of SSNA was also significantly elevated by whole body heating, which was indicated by the significant increase in the total autospectral power. Moreover, during whole body heating, the relative increase in SSNA spectral power was greater in the VHF region than in the LF and HF regions (Fig. 3). As observed in the present study (Fig. 1), Bini et al. (1) showed that, in the warm condition, SSNA bursts become narrow and intervals between bursts become shorter. Narrower SSNA bursts with shorter intervals between bursts are consistent with an increase in the spectral power of SSNA within the VHF region observed in the present study.
Cooling induced SSNA activation. The significant decrease in skin blood flow with elevated blood pressure during cooling indicates that cutaneous vasoconstrictor activity was increased by cooling. Although the spectral power of SSNA variability in the specified regions was elevated by cooling, the power distribution within these frequency regions during cooling was not significantly different from that during normothermia (Fig. 4B).
The difference in SSNA spectral power distribution between heat stress and normothermia/cooling is likely reflective of modulation of sudomotor/vasodilator vs. vasoconstrictor activities. During normothermia and cooling, SSNA is primarily vasoconstrictor in origin (1). Consistent with this observation, there was no significant difference in the power distribution between normothermia and cooling. A previous study showed maximum cutaneous vasoconstriction in response to external nerve stimulation in humans at 0.075–0.10 Hz (17). Those data may indicate that transmission from vasoconstrictor nerve to cutaneous vascular smooth muscle is more efficient within this frequency range (17). This might be the reason for the absence of a shift in the spectral distribution to the VHF region during cooling, because such a shift may reduce cutaneous vasoconstrictor responsiveness on the basis of the cited findings. Under the heat-stressed condition, sudomotor/vasodilator activities are the primary components of the SSNA signal (1, 10). The present data indicate a significantly greater increase during heat stress in spectral power in the VHF region than in the LF and HF regions (Fig. 3). Although the notion is speculative, it is possible that the greater power across the entire spectrum and, particularly, greater relative increase in VHF spectral power observed in this study are directly related to the temporal firing patterns that control eccrine sweat glands and/or cutaneous active vasodilation.
Consistent with previous observations (3, 5), heat stress reduced R-R interval and blood pressure variability within the LF and HF ranges and increased the LF/HF components of R-R interval variability. One reason for the decreased blood pressure variability during heat stress could be that increased vascular capacitance within the skin buffers the magnitude of the fluctuations in systemic blood pressure.
Cogliati et al. (2) reported significant coherence of LF and HF oscillations of SSNA with the corresponding components of R-R interval or blood pressure variability in normothermia in most of their subjects. In the present study, although integrated SSNA showed high coherence (K2 > 0.5) with R-R interval or blood pressure in a few subjects, coherence between these variables was low (K2 < 0.5) in most subjects across all frequency regions, regardless of the thermal condition (Table 3). Justification for differences in coherence between the present study and the study of Cogliati et al. is not clear but may be related to the aforementioned different algorithms used for the spectral analysis (i.e., the autoregressive model method vs. the Welch method). For the present data, such a low coherence may reflect control mechanisms for SSNA that are different from those for cardiovascular parameters (i.e., blood pressure and R-R interval). As discussed above, SSNA is primarily modulated by sk, Tco, and emotional stimuli. Numerous data suggest that SSNA is not controlled by baroreflexes, regardless of the thermal state of the individual (4, 6, 23, 24, 26, 27), whereas muscle sympathetic nerve activity, blood pressure, and heart rate are under strong baroreflex control. These observations confirm that SSNA is controlled by mechanisms different from those that control recognized baroreflex-modulated efferent responses. The lack of coherence between SSNA and blood pressure or R-R interval spectral variability supports this concept.
An early study (7) reported a loose coupling between spontaneous SSNA and resting respiratory rate in some subjects. In the present study, SSNA exhibited some oscillations within the HF range, which could be associated with respiratory rate (Fig. 2). Coupling of SSNA with respiration has been suggested to occur via reflex of respiratory muscle stretch or central effect of ventilation on SSNA (1, 2). However, given the relatively low level of SSNA under normothermic conditions, it is unlikely that spontaneous respiration greatly affected the frequency of SSNA. Thus SSNA variability showed low coherence with spontaneous respiration in most subjects in normothermia. On the other hand, whole body heating or cooling induced substantial increases in SSNA. Nevertheless, coherence between SSNA and respiration remained low in these thermal conditions. Together, these data challenge whether there is a close relation between respiration and SSNA, regardless of the level of SSNA. However, it should be emphasized that respiration was not controlled in the present study, and it is possible that coherence between SSNA and respiration may have been greater if respiration had been controlled at a fixed frequency.
Perspectives
SSNA spectral power and its distribution may represent a useful index for quantitative description of this neural signal. Typically, SSNA is expressed as total activity (i.e., burst area), as was used in the present study to show increases in SSNA with heat stress or cooling (Table 1). However, when expressed solely as total activity, this value does not identify whether changes in SSNA activation occurred via increase in amplitude and/or increase in burst frequency. Although quantification of SSNA via burst rate has been used in some studies (9, 11, 12, 16), SSNA burst number can be very difficult to count, because integrated SSNA bursts are irregular and are not synchronized with the cardiac cycle (7), and some bursts have more than one peak (Fig. 1). Moreover, burst rate does not distinguish whether the bursts are wide or narrow. Spectral power distribution of muscle sympathetic nerve activity (15) is recognized as an additional quantitative index of this neural activity. Spectral power distribution of the integrated skin sympathetic neural signal may provide another quantitative assessment of neural control of the skin.
In conclusion, the present data showed that the spectral power of SSNA variability distributes throughout the region of 0–2.5 Hz. Whole body heating and cooling significantly increased integrated SSNA spectral power; however, the greatest increase occurred in the VHF region during heating, but not during cooling. These data indicate that spectral distribution of SSNA during heat stress is different from that during normothermia and cooling, which may reflect differences in central modulation of sudomotor/vasodilator relative to vasoconstrictor neural activities.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Present address of J. Cui: Div. of Cardiology, Penn State College of Medicine, Hershey, PA 17033.
|
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 |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. A. Low, J. E. Wingo, D. M. Keller, S. L. Davis, R. Zhang, and C. G. Crandall Cerebrovascular responsiveness to steady-state changes in end-tidal CO2 during passive heat stress J Appl Physiol, April 1, 2008; 104(4): 976 - 981. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Yamazaki, K. Takahara, R. Sone, and J. M. Johnson Influence of hyperoxia on skin vasomotor control in normothermic and heat-stressed humans J Appl Physiol, December 1, 2007; 103(6): 2026 - 2033. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. E. Wilson, C. L. Sauder, M. L. Kearney, N. T. Kuipers, U. A. Leuenberger, K. D. Monahan, and C. A. Ray Skin-surface cooling elicits peripheral and visceral vasoconstriction in humans J Appl Physiol, October 1, 2007; 103(4): 1257 - 1262. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
, Normothermic baseline. Values are means ± SE (n = 7).