Gender affects sympathetic and hemodynamic response to postural stress

J. Kevin Shoemaker, Cynthia S. Hogeman, Mazhar Khan, Derek S. Kimmerly, Lawrence I. Sinoway

Abstract

We tested the hypothesis that differences in sympathetic reflex responses to head-up tilt (HUT) between males (n = 9) and females (n = 8) were associated with decrements in postural vasomotor responses in women. Muscle sympathetic nerve activity (MSNA; microneurography), heart rate, stroke volume (SV; Doppler), and blood pressure (Finapres) were measured during a progressive HUT protocol (5 min at each of supine, 20°, 40°, and 60°). MSNA and hemodynamic responses were also measured during the cold pressor test (CPT) to examine nonbaroreflex neurovascular control. SV was normalized to body surface area (SVi) to calculate the index of cardiac output (Qi), and total peripheral resistance (TPR). During HUT, heart rate increased more in females versus males (P < 0.001) and SVi and Qi decreased similarly in both groups. Mean arterial pressure (MAP) increased to a lesser extent in females versus males in the HUT (P < 0.01) but increases in TPR during HUT were similar. MSNA burst frequency was lower in females versus males in supine (P < 0.03) but increased similarly during HUT. Average amplitude/burst increased in 60° HUT for males but not females. Both males and females demonstrated an increase in MAP as well as MSNA burst frequency, mean burst amplitude, and total MSNA during the CPT. However, compared with females, males demonstrated a greater neural response (ΔTotal MSNA) due to a larger increase in mean burst amplitude (P < 0.05). Therefore, these data point to gender-specific autonomic responses to cardiovascular stress. The different MSNA response to postural stress between genders may contribute importantly to decrements in blood pressure control during HUT in females.

  • blood pressure
  • circulation
  • nervous system
  • vasculature

most reports(5, 10, 23, 32, 40), but not all (9), indicate that susceptibility for orthostatic intolerance is more common in women than in men. However, the mechanisms of this gender difference are unclear. The upright posture causes reductions in venous return leading to diminished stroke volume (SV) and cardiac output (Qi) (28). To maintain blood pressure, the reduction in SV must be compensated for by reflex mechanisms that increase total peripheral resistance (TPR). Otherwise, blood pressure falls, leading to cerebral hypoperfusion and syncope. Many reports (1, 2, 11, 24, 29, 41, 42) over several decades support the hypothesis that adequate sympathetic constrictor responses during the assumption of upright posture are critical for the maintenance of arterial pressure and cerebral perfusion. Whether orthostatic sympathetic vascular control is different between genders has not been investigated directly.

Gender differences in reflex-mediated sympathetic activation have only recently attracted investigative attention (5, 14, 17,22). However, these studies have focused on neural responses to nonbaroreflex stimuli, such as isometric handgrip contractions (8, 17), cold stress, and mental effort (17). The results of these earlier studies are equivocal in terms of whether or not gender differences exist in sympathetic reactivity. Reports (3, 5) of lower plasma norepinephrine concentrations in women during orthostatic stress suggest that sympathetic outflow may also be diminished. However, plasma norepinephrine concentrations are not a precise indicator of sympathetic outflow, and more direct measures are required to determine gender differences in sympathetic control during postural stress.

Therefore, the purpose of the current study was to test the hypothesis that sympathetic adjustments to tilt are attenuated in women versus men leading to diminished blood pressure responses to head-up tilt (HUT). To examine this hypothesis, microneurographic techniques were used to obtain direct recordings of postganglionic (7) sympathetic nerve traffic directed to blood vessels in skeletal muscle (7, 39) during graded HUT. In addition, measures of Qi and blood pressure were gathered to examine systemic neurovascular control.

METHODS

Subjects

A total of 17 healthy, nonsmoking, and normally active individuals volunteered for the study. There were nine males (30 ± 11 yr, 176 ± 4 cm, and 80 ± 12 kg; means ± SD) and eight females (26 ± 6 yr, 161 ± 8 cm, and 64 ± 12 kg). All individuals provided signed consent to the test procedures that had been approved by the Institutional Review Board at The Milton S. Hershey Medical Center.

Experimental Protocol

Two experimental procedures were performed on these subjects no sooner than 3 h after a meal and 12 h after caffeine ingestion. After establishment of baseline measures, data were collected during a 5-min period of supine rest, followed by a progressive HUT procedure that included 5 min at each of 20, 40, and 60° of tilt. The subjects were then returned to the supine position. After ∼10 min of quiet rest, a second baseline period was obtained, followed by insertion of a hand into ice water (∼4°C) for 2 min. During HUT, the increase in sympathetic outflow is due primarily to baroreceptor unloading whereas the cold pressor test (CPT) elicits a nonspecific response (38, 43). With the use of these two tests, we examined whether gender differences in sympathetic reflex responses to HUT were present and if these were reflex specific.

Measurements

Heart rate was determined by standard electrocardiogram methods. Arterial pressure was measured continuously from the finger of the left hand by photoplethysmographic methods (model 2300 Finapres, Ohmeda; Englewood, CA). The hand from which blood pressure was obtained was maintained at heart level throughout the testing periods. Baseline blood pressures from the Finapres device were corrected against manually obtained systolic and diastolic measures before the onset of the data collection.

Sympathetic activation was assessed by microneurographic (37) measures of muscle sympathetic nerve activity (MSNA) in the common peroneal nerve. A 200-μm diameter, 35-mm long tungsten microelectrode that tapered to an uninsulated 1- to 5-μm tip was inserted transcutaneously into the peroneal nerve just posterior to the fibular head. A reference electrode was positioned subcutaneously 1–3 cm from the recording site. Neuronal activity was amplified 1,000 times by a preamplifier and 50–100 times by a variable gain isolated amplifier. The signal was band-pass filtered with a bandwidth of 700–2,000 Hz and then was rectified and integrated to obtain a mean voltage neurogram (0.1-s time constant). A MSNA site was confirmed by manually manipulating the microelectrode until the characteristic pulse-synchronous burst pattern was observed that did not produce skin paresthesias and that increased in frequency during a voluntary apnea but not during arousal to a loud noise (7).

Cardiac SV velocity and aortic dimensions were obtained to calculate Qi. Aortic diameter was obtained using two-dimensional B-Mode echo Doppler imagine (2.5-MHz probe) with a parasternal long-axis view of the aorta. Our preliminary studies have indicated that aorta diameters ranged from 2.7 ± 0.1 to 2.8 ± 0.1 cm between supine and 45° HUT (n = 6; not significant). Therefore, the aortic measurements for this study were made while the subjects were supine. SV velocity was obtained from the suprasternal notch using a hand-held 2-MHz pulsed wave probe (model 500 M, Multigon; Yonkers, NY).

Data Analysis

Analog signals for blood pressure, MSNA, and SV velocity, sampled at 200 Hz, and for the electrocardiogram, sampled at 400 Hz, were collected with an on-line data acquisition and analysis package (PowerLab, ADInstruments). Mean arterial pressure (MAP) was calculated as systolic blood pressure + 1/3 diastolic blood pressure. Pulse pressure (PP) was determined as the difference between systolic and diastolic blood pressures. SV was calculated as the product of SV velocity and the aortic cross-sectional area for the mean cardiac interval. To diminish concerns that gender differences in body mass and blood volume (5) might interfere with interpretation of gender differences in the hemodynamic responses SV was normalized to body surface area (SVi) for all subjects. Qiwas then calculated as HR · SVi. TPR was calculated as the quotient of MAP and Qi. Bursts of MSNA activity with a 2:1 or greater signal-to-noise ratio were measured for amplitude/burst and frequency/minute during the 5-min baseline segment and for the final 2–3 min of each HUT phase. The average burst frequency, burst incidence (bursts per 100 heartbeats), and mean burst amplitude were determined from these time segments. Total MSNA was calculated as the sum of the analog burst amplitudes obtained during each min of data analysis and averaged to provide a value of total amplitude/min. To examine the distribution characteristics of amplitude modulations with HUT, the median burst amplitude was determined from all bursts analyzed over the 5 min of baseline and each HUT phase after normalizing all bursts to the largest burst that was given a value of 1. For the CPT average hemodynamic and MSNA data were obtained over 5 min at rest and between 60 and 90 s after insertion of the hand into the ice bath.

Statistics

The effects of HUT and the CPT on MSNA and hemodynamic variables were analyzed using a repeated measures analysis of variance with subjects grouped by gender. Tukey's post-hoc analysis was performed to estimate differences among means. Probability levels during multiple pointwise comparisons were corrected using Bonferroni's approach. Within-group differences in the median burst amplitude between supine and HUT were assessed using a two-tailed paired t-test with Bonferroni's correction for multiple comparisons. Comparisons of sympathetic responses to 60° HUT and the end CPT (i.e., the changes from baseline in burst frequency, incidence and amplitude and total MSNA) were examined using a two-tailed paired t-test. The level of statistical probability was set at P < 0.05. Values are presented as means ± SE.

RESULTS

Head-Up Tilt

Hemodynamics.

Baseline levels of heart rate, MAP, PP, SVi, Qi, and TPR were not different in men and women (Table1). Heart rate and MAP increased in both groups during HUT (Fig. 1). Significant group · tilt interactions were observed for heart rate and MAP (P < 0.002) and PP (P < 0.004). Specifically, heart rate increased more in the females than it did in the males (P < 0.05; Fig. 1), and MAP increased more in the males than in the females at 60° HUT (Fig. 1). PP at 60° HUT was greater in the males (67 ± 4 mmHg) than in the females (52 ± 4 mmHg) (P < 0.006). A gender · tilt interaction was also present for the effect of tilt on PP (i.e., ΔPP) (P < 0.002). Specifically, at 60° HUT the ΔPP was −12 ± 3.8 and 2.06 ± 3.34 mmHg for women and men, respectively (P < 0.001). SV and Qi decreased similarly in both groups during HUT. At the highest tilt angle, SVi decreased 36 ± 7 ml/m2 in the males and 47 ± 4 ml/m2 in the females (P = 0.11). The reduction in Qiat 60° HUT was 1.6 ± 0.5 l/min in the males and 2.0 ± 0.2 l/min in the females (P = 0.31; Fig. 1). TPR increased similarly in both groups (Fig. 1).

View this table:
Table 1.

Baseline hemodynamic values before head-up tilt

Fig. 1.

Effect of gender on the hemodynamic responses between supine and progressive increases in head-up tilt angle. ΔHR, change in heart rate; ΔMAP, change in mean arterial pressure; ΔSVi, change in stroke volume normalized to body mass; ΔQi, change in cardiac output; ΔTPR, change in total peripheral resistance. *P < 0.05, Probability vs. baseline; α P < 0.05, pointwise gender effect.

MSNA.

Baseline MSNA burst frequency was 6.8 ± 1.6 beats/min in females and 16 ± 2 beats/min in males (P < 0.07). Baseline burst frequency normalized to heart rate (burst incidence) was higher in the males (26 ± 4 beats/100 heartbeats) than the females (11 ± 2 beats/100 heartbeats; P < 0.05; Fig. 2). The baseline differences led to a main effect of group (P < 0.03) with males demonstrating a consistently higher burst frequency during HUT.

Fig. 2.

Effect of gender on the sympathetic neural characteristics between supine and progressive increases in head-up tilt angle. Freq, burst frequency; Amp/burst, mean amplitude per burst; au, arbitrary units. *P < 0.05, Probability vs. baseline; α P < 0.05, pointwise gender effect.

For the females, mean burst amplitude at 60° HUT [30 ± 3 arbitrary units (au)] was not different from baseline (30 ± 6 au). For the males, burst amplitude increased from 29 ± 4 au at rest to 40 ± 4 au at 60° (P < 0.01; Fig. 2). The effect of HUT and gender on the burst amplitude was also reflected in the analysis of median burst amplitude. For the males, the median burst amplitude (after normalizing to the largest burst) increased from 0.35 ± 0.03 units in supine to 0.43 ± 0.03 (P = 0.16), 0.44 ± 0.05 (P = 0.09), and 0.49 ± 0.03 (P < 0.009) units during 20, 40, and 60° HUT, respectively. In contrast, median burst amplitude for the females was not significantly changed between supine (0.37 ± 0.03 units) and 20° (0.36 ± 0.03 units;P = 0.82), 40° (0.41 ± 0.04 units;P = 0.25), and 60° (0.43 ± 0.05 units;P = 0.21). As a result of the similar increases in burst frequency in both groups but greater change in burst amplitude in the males, total amplitude at 60° HUT was higher in males (1,699 ± 245 au) compared with the females (1,026 ± 184 au) (main effect of group, P < 0.05; Fig. 2). The same conclusion is reached when the MSNA response to 60° HUT (i.e., ΔMean Amp and ΔTotal Amp) is assessed (Fig.3). Specifically, as the change in burst frequency was similar in both groups it was the lack of change in burst amplitude that contributed to the smaller ΔTotal Amp in the females.

Fig. 3.

Effect of gender on the changes in muscle sympathetic nerve activity (MSNA) from baseline as measured at 60° of head-up tilt. A: change in burst frequency. B: change in Amp/burst. C: change in total MSNA amplitude. *P < 0.05, Probability vs. male.

Cold Pressor Test

Hemodynamics.

Heart rate and Qi were not altered during the CPT in either male or female groups (Fig. 4). MAP and TPR increased similarly during the CPT in both males and females (Fig.4).

Fig. 4.

Effect of gender on the hemodynamic responses to a cold pressor test. HR (A), Qi normalized to body mass (B), MAP (C), and TPR (D). *P < 0.05, Probability vs. baseline.

MSNA.

Compared with baseline, MSNA burst frequency and burst incidence increased during the CPT (P < 0.05) in both men and women (Fig. 4). Unlike the HUT test, mean burst amplitude increased in both females (from 25 ± 4 to 33 ± 6 units;P < 0.02) and males (from 40 ± 4 to 58 ± 10 units; P < 0.002) between baseline and CPT (Fig.5). However, compared with the females, the total MSNA response (ΔTotal Amp) to the CPT was greater (P < 0.05) in the males due to a similar ΔFrequency but greater ΔMean burst amplitude (P < 0.05; Fig.6). Thus the different response to a CPT in females versus males is qualitatively similar to that for HUT.

Fig. 5.

Effect of gender on the sympathetic neural responses to a cold pressor test. Freq (A), burst incidence (B), Amp/burst (C); and total amplitude per minute (D). *P < 0.05, Probability vs. baseline.

Fig. 6.

Effect of gender on the changes in sympathetic nerve activity to a cold pressor test. ΔFreq (A), ΔAmp/burst (B), and ΔTotal Amp/burst (C). *P < 0.05, Probability vs. male response.

DISCUSSION

There is debate as to whether or not gender differences exist in sympathetic reflex responses; most (8, 34) but not all (17) reports suggest that males demonstrate greater cardiovascular responses to various stressors. In the current study, the integrated cardiovascular and sympathetic adjustments to graded HUT were investigated in men and women to examine gender differences in orthostatic reflex responses. The primary new finding was the smaller increase in both blood pressure and sympathetic outflow in females versus males during the higher levels of HUT (60°). The attenuated MSNA response in the females was related to a reduced ability to augment sympathetic burst amplitude. To examine the reflex-specificity of this response, the hemodynamic and neural responses to a CPT were also assessed. There was a smaller overall increase in total MSNA in the females during this test as well. Although baseline differences between groups contributed to absolute differences in MSNA levels during the test protocols, the diminished neural responses in the females were due to attenuated changes in burst amplitude and not burst frequency. Interestingly, unlike the HUT response, women did demonstrate an ability to augment sympathetic burst amplitude during the CPT, but to a smaller magnitude than the male counterparts. Therefore, these data point to a generalized gender difference in sympathetic reflex responsiveness.

Despite evidence of greater orthostatic intolerance in women (5,10, 23, 32, 40) and of the important role that sympathetic neurovascular control plays in maintaining blood pressure during upright posture (12, 24), little is known of sympathetic reflex responses to HUT between genders. The lower levels of baseline MSNA burst frequency in the females observed in the current study confirm earlier reports (17, 25, 35). Because of the lower baseline levels the absolute burst frequency was less for the females at each level of HUT but the expected linear response with tilt angle (15) was maintained and the increase above baseline was the same in both groups. However, the total amplitude response was diminished in the females due to an inability to increase mean burst amplitude. This finding suggests that the earlier reports of diminished plasma norepinephrine concentrations during HUT (3) and lower body suction (5) in women were due, at least in part, to attenuated postganglionic sympathetic nerve activity.

In the current study, total MSNA during HUT was diminished in the females despite lower MAP and smaller pulse pressures. The smaller sympathetic outflow despite greater stimuli suggests that important differences exist in either sensory afferent sensitivity or in the central integration and modification of reflex input signals. Sensory signals contributing to the HUT sympathetic response arise from cardiac chambers, aortic and carotid structures, and from vestibular receptors. Otolithic vestibular afferents appear to contribute to the orthostatic sympathetic response (44). However, available evidence (27) suggests that gender does not affect vestibulosympathetic activation.

To our knowledge, information on gender differences in regional baroreceptor function is not well defined. Afferent neurons in the cardiac chambers and pulmonary structures detect changes in venous return exerting control over limb sympathetic nerve activity (16). The absolute level of volume shift during upright posture may be less in women than men (23) but smaller body mass and blood volume suggest that reductions in cardiac filling pressures may not be different. The similar reductions in normalized SV and Qi in the current study support the proposition that venous filling pressures were not a dominant factor in determining the gender-specific autonomic response to HUT.

Aortic arch baroreceptors exert powerful inhibitory signals for MSNA during increases in MAP (31). A greater sensitivity of these receptors in females to increases in MAP would result in augmented sympathoinhibition. However, the rise in MAP during HUT for the females was significantly less than in males discounting a possible role of greater aortic baroreflex gain.

Hydrostatic gradients with HUT reduce the distending pressure in the carotid sinus resulting in a sympathoexcitatory stimulus. This reduction in carotid sinus distending pressure would partially be corrected by the increase in MAP (measured at heart level). Therefore, the greater reduction in PP in the women compared with men should have provided a greater baroreflex-mediated increase in sympathetic outflow in the females but this did not occur.

Therefore, these data support the hypothesis that baroreflex sensitivity for control of sympathetic outflow is diminished in women. The details regarding the effect of gender on baroreflex sympathetic control remain unclear (14).

The observation that changes in burst frequency were the primary factor contributing to increased total MSNA at 20° and 40° of HUT is in agreement with earlier reports of autonomic responses to graded HUT (4, 6) and lower body negative pressure (33). The increase in burst amplitude is probably the combined function of multiple discharges from single neurons during a cardiac cycle and/or to a larger number of neurons contributing to the integrated neurogram (21, 26). Therefore, the lack of increase in mean burst amplitude for the females suggests that important gender differences exist in central neural recruitment patterns in response to postural stress. This conclusion is strengthened by the current observation that while burst amplitude was not increased during HUT in the females, it was increased during a CPT. However, the CPT may be a more potent stimulus for changes in MSNA burst amplitude than HUT. Also, it is acknowledged that the role of changes in burst frequency versus amplitude in determining the vasomotor response is not known. Our observation that mean burst amplitude was not increased in the women during HUT but was during the CPT, albeit to a lower level than the males, suggests that specific tilt-induced reflex sympathetic control may be diminished in women.

In contrast to the MSNA response, women demonstrated a greater reflex increase in heart rate during HUT than the males. This latter observation is consistent with previous reports of heart rate responses during orthostatic stress (3, 5). In addition, the greater tilt-induced heart rate response in females (5) supports conclusions based on spectral analysis of heart rate dynamics that women have greater vagal cardiac control at rest and, therefore, greater parasympathetic withdrawal during HUT (3).

It was expected that the smaller MSNA response to HUT in women would be associated with a smaller increase in TPR compared with the male response. However, this was not the case. This is particularly interesting because the decrements in Qi were similar in males and females but MAP increased more in the males. It is noteworthy that the same results are observed whether the constrictor response to HUT is expressed as TPR or total vascular conductance. This apparent dissociation between MSNA and TPR must be due to additional factors that contribute to blood pressure stability. In addition to sympathetic nerve activity the vasoconstrictor response is dependent upon the levels of neurotransmitter release, adrenergic receptor density and postsynaptic mechanisms that control vascular smooth muscle calcium levels. Currently, conclusions on gender effects on each of these factors are unclear. There is evidence both for (20) and against (5) greater adrenergic receptor sensitivity in women. As far as we are aware there are no data relating the effect of gender to norepinephrine spillover. Moreover, during HUT nonsympathetic vasoconstrictor mechanisms such as local myogenic or veno-arteriolar constrictor responses (13, 18, 19) are likely to contribute importantly to the overall change in TPR as well. Furthermore, it is not known how regional vascular responses to sympathetic excitation vary between males and females. As such, these data suggest that blood pressure control during HUT differs in healthy men and women.

Limitations

In this study testing was not constrained to a particular menstrual phase in the female group. Menstrual phase appears to affect the neural responses to sympathetic stimulation during exercise (8, 22). On the basis of very low frequency blood pressure oscillations, some have argued that sympathetic responses to HUT are also modified by the menstrual phase (30). However, direct comparisons of MSNA and frequency-specific spectral power characteristics of heart rate and blood pressure have not supported the conclusion that low-frequency power is indicative of sympathetic modulations (35). Therefore, menstrual phase effects on baroreflex sympathetic control remain uncertain. In addition, three of the eight female volunteers in this study were on oral contraceptives whereas the other five were not adding additional variability to the hormonal profile of these women. However, assuming that these women were not on the same temporal cycle, the current data may be representative of the generalized female response pointing to an important gender difference in the integrated cardiovascular and neurovascular orthostatic response that may only be modified by menstrual phase.

A second limitation of this study is that blood volume is likely to be smaller in women (5). This variable was not measured in the current study. However, it is unlikely that this factor interfered with the current observations. For example, Thompson et al. (36) observed that hypovolemia augmented the vasoconstrictor response to lower body suction suggesting that the corresponding sympathetic response was augmented. Therefore, smaller blood volume per se in females is unlikely to explain smaller MSNA responses to HUT. Regardless, SV was normalized to the body surface area to account for this possible limitation.

In summary, upright posture imposes cardiovascular stress through the translocation of 500–800 ml of blood from the thoracic circulation to the capacitance vessels of the viscera and legs. Appropriate autonomic responses that affect peripheral vascular tone and heart rate are required to minimize the effects of upright posture on venous return and Qi. The current study provides a physiological basis for the common finding that women are more susceptible to orthostatic hypotension than men. Specifically, lower MSNA responses in women compared with men were associated with smaller increases in blood pressure during the graded HUT stress. The diminished sympathetic response may be protective for women in terms of vascular disease (34) but it may increase the susceptibility of women for orthostatic hypotension.

Acknowledgments

We are grateful to M. Herr and S. Quraishi for expert technical assistance during data collection.

Footnotes

  • This study was supported by Cooperative Activities Program, Natural Sciences and Research Council of Canada (NSERC) and Canadian Space Agency Grant 216758-98, by NSERC Grant 217916-99 (to K. Shoemaker), by National Aeronautics and Space Administration Grant NAG-9-1044, by a Veterans Administration Merit Review Award, by National Institute on Aging Grants R01-AG-12227, and by National Heart, Lung, and Blood Institute Grant K24-HL-04011 (all to L. I. Sinoway). This study was also sponsored by National Center for Research Resources Grant M01-RR10732.

  • Address for reprint requests and other correspondence: J. K. Shoemaker, Neurovascular Research Laboratory, School of Kinesiology, Rm. 3110, Thames Hall, Univ. of Western Ontario, London, Ontario, Canada N6A 3K7 (E-mail: kshoemak{at}uwo.ca).

  • 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

View Abstract