INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ORTHOSTATIC INTOLERANCE is commonly seen after a period of prolonged bed rest (9), as well as in astronauts after their return to Earth (22). Although some of the mechanisms contributing to postspaceflight orthostatic intolerance have been identified, their relative importance is unknown, and additional mechanisms may also be involved. Demonstrated and possible contributing factors include 1) hypovolemia, which contributes to a decrease in stroke volume and increase in heart rate on reentry into the gravitational field (22), 2) increased lower body venous compliance perhaps due in part to muscle atrophy and reduction of venous smooth muscle tone (22), which would augment pooling of blood in gravity, 3) decreased or insufficient systemic sympathoexcitatory (10) and vasoconstrictor (3, 10, 14) responses to orthostasis, compromising the body's ability to maintain arterial blood pressure, and 4) possible decreased cerebral vascular compliance as an adjustment to the increased arterial and intracranial pressures during 0 g, which may jeopardize 1 g orthostatic cerebral perfusion (22).

Although significant differences in heart rate and stroke volume responses to orthostasis were not found between orthostatic-tolerant and intolerant astronauts after flight, insufficient vasoconstrictor responses appeared only in individuals unable to tolerate 10 min of standing after flight (3). Another study also found that postflight total peripheral resistance responses to orthostasis had decreased in comparison to preflight responses, although this study did not mention whether these astronauts tolerated the orthostasis (14). Whitson et al. (23) found no significant change in total peripheral resistance from before to after flight (including data from orthostatically tolerant astronauts); however, postflight plasma catecholamine levels were significantly increased relative to preflight levels, suggesting that the peripheral vascular response to sympathetic nerve activity (SNA) had been decreased by microgravity. Fritsch-Yelle et al. (10) found that astronauts who could not tolerate 10 min of orthostasis on landing day had significantly decreased standing peripheral vascular resistance and significantly smaller increases in standing plasma norepinephrine levels than astronauts who could tolerate the 10 min of orthostasis.

Because astronauts often experience substantial sleep restriction and interruption (11, 19), we hypothesized that this sleep restriction can lead to attenuated vasoconstrictor responses to orthostasis and possibly other changes in autonomic control, thereby compromising the ability to maintain blood pressure and contributing to orthostatic intolerance. To test this hypothesis, we studied human subjects before and after 4 nights of sleep restriction (4 h total sleep/night) to determine the effect of sleep restriction on baroreflex function, cardiovascular responses to lower body negative pressure (LBNP), LBNP tolerance, and heart rate variability (a generally accepted measure of vagal activity).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study was approved by the University of North Texas Health Science Center Institutional Review Board. Ten healthy volunteers (4 women and 6 men, ages 22-46 yr) participated in the study after giving written, informed consent. All subjects completed a medical history-health questionnaire and passed a physical examination. The subjects reported sleeping 7.7 h/night on average (range 6.3-9.5 h), with occasional naps of 0.3- to 1.4-h duration. Subjects were not taking any medication routinely (prescription or over-the-counter). Female subjects tested negative for pregnancy and were not studied during or within 2 days of menses to eliminate potential confounding effects of menses on fluid metabolism, blood volume, and cardiovascular function. During the study period, subjects maintained their normal, pre-sleep-restriction caffeine intake and abstained from alcohol as well as vigorous exercise. Moderate exercise consistent with the subjects' normal routine was allowed.

Measurements

During the experimental protocols the following measurements were obtained. Heart rate was obtained using a standard limb-lead ECG. Arterial blood pressure was measured noninvasively from beat-to-beat photoplethysmographic recordings at the finger (Finapres blood pressure monitor 2300, Ohmeda, Englewood, CO). Noninvasive pulsed Doppler flow velocity measured at the brachial artery was used with two-dimensional ultrasonically measured artery cross-sectional area to calculate forearm blood flow (Interspec XL, Conshohocken, PA, presently owned by ATL, Bothell, WA). This technique has been commonly utilized in previous studies and validated against occlusion plethysmography (4, 12, 20). Briefly, brachial arterial blood flow velocity was quantified for 10 cardiac cycles during each measurement time period. Brachial artery cross-sectional area was measured at the probe site and quantified for each measurement time by built-in software. Forearm blood flow was then calculated as the product of velocity and vessel area. Forearm vascular resistance was calculated as mean arterial pressure divided by forearm blood flow.

Protocol

Baseline cardiovascular testing was performed on the day after baseline sleep monitoring, and post-sleep-restriction cardiovascular testing was performed on the day after the 4th night of sleep restriction. Subjects were tested during a morning or an afternoon session, but the time of testing for each individual was maintained between baseline and post-sleep-restriction sessions. Subjects were instrumented for recording of ECG, arterial blood pressure, and forearm blood flow. ECG and arterial blood pressure data were stored to an IBM-compatible Pentium-based computer with utilization of WINDAQ data acquisition hardware and software (Dataq Instruments, Akron, OH), and forearm blood flow data were recorded to VHS tape. A catheter was also inserted for drug administration during the arterial baroreflex tests. All tests were performed with the subjects supine.

Arterial baroreflex function was determined from blood pressure and heart rate responses to injections of nitroprusside followed by phenylephrine (6, 7). Phenylephrine was administered at the nadir in systolic pressure caused by the nitroprusside. Sodium nitroprusside and phenylephrine were given in increasing doses until a 10- to 20-mmHg change in systolic pressure from baseline was seen in both directions. The dose combination producing the desired response was then repeated a second time. Subjects were closely monitored to ensure a return to baseline heart rates and blood pressures between trials.

Orthostatic tolerance was determined with LBNP. Subjects were placed in an airtight chamber enclosing them from the waist (superior iliac crests) down. A vacuum cleaner attached to the chamber created the negative pressures. Subjects were exposed to progressively more negative pressures: 10, 20, 40, 50, and 60 mmHg. Each pressure was maintained for 3 min, except 60 mmHg, which was maintained for the duration of the subject's tolerance, with a limit of 18 min at 60 mmHg. The test was terminated when 1) the subject's blood pressure began to fall significantly and recovery without test termination was not anticipated, 2) the subject reported symptoms of nausea or dizziness, or 3) the subject tolerated 60 mmHg for 18 min. The test was preceded by 1 min of recorded baseline and followed by 1 min of recorded recovery.

Heart rate variability was assessed during baseline conditions and during LBNP (50 or 60 mmHg). Baseline data collection consisted of the subject breathing to a metronome at 15 breaths/min for 5 min. Subjects were asked to breathe at a relatively constant tidal volume, representative of normal respiration under quiet, restful conditions. Baseline data were collected before the arterial baroreflex and LBNP tests. Heart rate variability was also assessed during 60-mmHg LBNP, unless the subject did not tolerate 60 mmHg for a reasonable number of heartbeats, in which case 50-mmHg data were used.

Analyses

Arterial baroreflex. We quantified arterial baroreflex function by determining the change in R-R interval and heart rate for a given change in systolic arterial blood pressure during the drug-induced rise in blood pressure, a method that has been utilized by previous investigators (17, 21). Arterial blood pressure pulses were matched with a following R-R interval presumed to be affected by that blood pressure pulse, as described previously (16).

LBNP. Orthostatic tolerance was calculated as the sum of the products of each level of negative pressure multiplied by the duration of toleration of that pressure. Responses to LBNP were assessed by determining changes from baseline in systolic, diastolic, and mean arterial blood pressure, heart rate, forearm blood flow, and forearm vascular resistance for each level of negative pressure. Data were analyzed during the last 30 s at each pressure level, except 60 mmHg, for which data were analyzed during the last 30 s of every 3-min interval as well as the final 30 s of tolerance. Baseline and recovery values were obtained from 1-min data segments. Two-factor repeated-measures ANOVA determined whether significant differences existed between experimental conditions for each variable (factors: LBNP level and sleep restriction status). Post hoc paired t-tests were then used to delineate which specific changes were significant.

Heart rate variability. Heart rate variability estimated in the time domain was used to determine whether sleep restriction caused changes in basal parasympathetic activity levels as well as changes in autonomic responses to a physiological stress (LBNP in this case). Baseline data, as well as data collected during 60-mmHg LBNP, were used in the comparison between before and after sleep restriction. R-R intervals from LBNP data were taken only after subjects had equilibrated to the negative pressure but well before any presyncopal symptoms developed. Equal numbers of R-R intervals were analyzed within each subject, with a minimum of 138 R-R intervals analyzed. Heart rate variability was quantified by calculating the root mean square of sequential differences of R-R intervals. As described above, two-factor repeated-measures ANOVA and post hoc paired t-tests determined effects of sleep restriction and LBNP on heart rate variability. All statistical analyses were performed at a significance level of 0.05. Values are means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sleep Architecture

Arterial Baroreflex Function

1

1

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Comparison of baroreflex gain (n = 6), lower body negative pressure (LBNP) tolerance (n = 10), and heart rate variability (n = 9) before (pre-SR) and after (post-SR) 4 nights of sleep restriction. Heart rate variability values during LBNP are at 60 mmHg.

Orthostatic Tolerance

The normal physiological response to orthostasis was observed before and after sleep restriction. LBNP significantly affected all dependent variables within both sleep conditions as follows: systolic, diastolic, and mean arterial pressures decreased with increasing LBNP (ANOVA: P < 0.001, P = 0.002, P < 0.001, respectively; Fig. 2, Table 1); heart rate increased with increasing LBNP (ANOVA: P < 0.001; Fig. 3); forearm vascular resistance increased with increasing LBNP (ANOVA: P < 0.001; Fig. 4); and forearm blood flow decreased with increasing LBNP (ANOVA: P < 0.001; Table 1).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Effect of sleep restriction on systolic blood pressure responses to LBNP. * Significantly greater than before sleep restriction (P < 0.05). # Significantly below baseline within a sleep condition (P < 0.05). Values at 60 mmHg are from last 30 s of first 3 min spent at 60 mmHg. One subject did not reach 60 mmHg before sleep restriction, so this data point is average from remaining 9 subjects. Data points for "last 30 s" correspond to average of final 30 s of subjects' LBNP tolerance. Values are means ± SE.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Effect of sleep restriction on heart rate responses to LBNP. * Significantly less than before sleep restriction (P < 0.05). # Significantly above baseline within a sleep condition (P < 0.05). Values at 60 mmHg are from last 30 s of first 3 min spent at 60 mmHg. One subject did not reach 60 mmHg before sleep restriction, so this data point is average from remaining 9 subjects. Data points for "last 30 s" correspond to average of final 30 s of subjects' LBNP tolerance. Values are means ± SE.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Effect of sleep restriction on forearm vascular resistance responses to LBNP. # Significantly above baseline within a sleep condition (P < 0.05). Values at 60 mmHg are from last 30 s of first 3 min spent at 60 mmHg. One subject did not reach 60 mmHg before sleep restriction, so this data point is average from remaining 9 subjects. Data points for "last 30 s" correspond to average of final 30 s of subjects' LBNP tolerance. Values are means ± SE.

In contrast to our hypothesis, sleep restriction did not impair the physiological responses to LBNP. Systolic pressure tended to be greater at all levels of LBNP after sleep restriction, and this trend was enhanced at the more severe levels of LBNP (Fig. 2). The relative increase in systolic pressure after sleep restriction was statistically significant during the first 3 min at 60 mmHg (110 ± 6 and 124 ± 3 mmHg before and after sleep restriction, respectively, P = 0.038) and near significance during the last 30 s of tolerance (97 ± 8 and 111 ± 6 mmHg before and after sleep restriction, respectively, P = 0.053). Although sleep restriction did not significantly affect diastolic and mean arterial pressures, both showed the same trend to be higher after sleep restriction, and this trend was also enhanced at the more severe levels of LBNP (Table 1).

Heart rate tended to be lower at all levels of LBNP after sleep restriction, and this relative reduction became more apparent at the more severe levels of LBNP (Fig. 3). During the last 30 s of LBNP, heart rate was significantly lower after sleep restriction than in sleep-replete conditions (108 ± 8 and 99 ± 8 beats/min before and after sleep restriction, respectively, P = 0.028).

Although neither forearm blood flow nor forearm vascular resistance responses to LBNP were significantly affected by sleep restriction, one difference between pre- and post-sleep-restriction responses was noted. When LBNP reached 60 mmHg before sleep restriction, forearm vascular resistance decreased to levels near pre-LBNP baseline and remained there for the balance of the test (Fig. 4). This decrease in forearm vascular resistance at 60 mmHg was reflected in forearm blood flow, which decreased continually during LBNP until 60 mmHg, at which time it increased slightly (Table 1). After sleep restriction, however, forearm vascular resistance remained above baseline levels throughout LBNP, exhibiting only a slight decline at the end of tolerance.

Heart Rate Variability

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The effect of sleep restriction on sleep architecture was consistent with the known effects of sleep restriction (1, 2). Although sleep restriction of 4 h/night for 4 consecutive nights did not affect orthostatic tolerance, the ability to maintain blood pressure in the face of an LBNP challenge appears to be equal to or augmented slightly after sleep restriction, inasmuch as LBNP-induced systolic blood pressure reduction and heart rate elevation at more stressful LBNP levels were each attenuated after sleep restriction. The lesser post-sleep-restriction heart rate response to LBNP may simply be a baroreflex-mediated response to the higher blood pressures maintained after sleep restriction. Sleep restriction also had no effect on arterial-cardiac baroreflex function or heart rate variability; therefore, sleep restriction for 4 nights does not appear to impair reflex control of cardiovascular function.

Arterial-Cardiac Baroreflex Function and Heart Rate Variability

A related study in our laboratory provided some insight into the changes in sympathetic responses to alternative stresses after sleep restriction. It examined the effects of sleep restriction on catecholamine responses to the combined physical and mental stress of handgrip exercise with simultaneous timed tasks employing logical, spatial, and mathematical skills. Results showed a trend toward decreased catecholamine release in response to stress after sleep restriction (unpublished observations). Nevertheless, in the present study the support of arterial pressure (perhaps mediated by increased systemic vascular resistance) was not impaired; therefore, our data suggest that sleep restriction did not adversely affect the control of parasympathetic or sympathetic activity.

Orthostatic Tolerance

Mean arterial pressure is determined by heart rate, stroke volume, and systemic vascular resistance, with the product of the former two composing cardiac output. Heart rate tended to be lower during LBNP after sleep restriction, so clearly the tendency for increased blood pressure was not a result of heart rate elevation. Increased systemic vasoconstrictor responses appear to participate in the improved blood pressure maintenance. Before sleep restriction, forearm vascular resistance dropped sharply when LBNP reached 60 mmHg and remained at levels not significantly different from baseline for the remainder of the orthostatic challenge. After sleep restriction, forearm vascular resistance remained above baseline when LBNP reached 60 mmHg and remained significantly greater than baseline levels for the duration of the orthostatic challenge. These changes in forearm vascular resistance responses may represent systemic resistance elevation responsible for the relatively greater systolic pressures observed at severe LBNP levels after sleep restriction.

Lusardi et al. (13) found increases in resting heart rate and systolic blood pressure after 1 night of sleep restriction (4.5 h), so contrary to our study, heart rate elevation may have played a role in the increased systolic pressure they observed. A possible explanation for the discrepancy is that their observed changes in cardiovascular parameters occurred shortly on awakening, and it is unclear whether their subjects were still lying down, sitting, or standing during these measurements. Our heart rate and blood pressure measurements were taken well after the subjects had awakened and while they had been in the supine position for 30 min. Another study agreed with our finding of a decreased heart rate, although the decrease occurred under resting as well as stressful conditions, and the stress in their case was dynamic exercise rather than simulated orthostasis (5). Also, that study employed 30 h of sleep deprivation as opposed to sleep restriction, possibly indicating that total sleep deprivation has a more pronounced effect than sleep restriction on heart rate, in that the effect is apparent even at rest.

Stroke volume is another determinant of arterial pressure via its contribution to cardiac output. We can only speculate as to the effect of sleep restriction on stroke volume; we did not measure this parameter. Blood volume, one of the determinants of stroke volume, does not appear to be affected by sleep restriction. Our subjects' weight and hematocrit, both good indicators of changes in body fluid volume over periods of days, were unchanged from before to after sleep restriction (mean weight: 76.4 and 76.5 kg before and after sleep restriction, respectively; mean hematocrit: 38.8 and 38.4% before and after sleep restriction, respectively). The decreased heart rate would increase ventricular filling time and, therefore, stroke volume, but it is doubtful that this increase would be sufficient to overcome the slower heart rate (which has a negative effect on cardiac output) and effectively raise cardiac output and, thus, arterial pressure. Myocardial contractility, a major determinant of stroke volume, was not measured; however, we would expect an increase in contractility to be accompanied by an increase in heart rate, inasmuch as both are stimulated by cardiac SNA. Because of the trend toward a decreased, rather than an increased, heart rate, we are confident that myocardial contractility did not increase. Therefore, it is unlikely that cardiac effects played a role in the trend toward increased arterial pressures observed after sleep restriction. It is important to note that we are drawing conclusions regarding myocardial contractility from heart rate data, yet the sympathetic branch of the autonomic nervous system is more important in the control of the former, whereas the parasympathetic branch is more important in the control of the latter, at least at resting heart rates. Thus our interpretations remain tentative.

It seems plausible that the trend toward better orthostatic blood pressure regulation after sleep restriction may result from a greater proportion of time spent upright during sleep restriction (20 vs. ~16 h/day under normal conditions). Subjects spent more time compensating for the effects of gravity on the circulation during sleep restriction, and perhaps this increased their ability to maintain arterial pressure during orthostatic stress. As mentioned above, a combined physical-mental exercise stress showed a trend toward a decreased release of catecholamines after sleep restriction. This observation, if true for an orthostatic stress as well and combined with the increased vasoconstrictor response at more severe LBNP levels, could suggest an increased sensitivity of the vasculature to sympathetic nerve activation after sleep restriction; however, we did not assess this aspect of vascular function in this study.

If the increased percentage of time spent counterbalancing gravity during sleep restriction does indeed improve blood pressure maintenance during orthostasis, this effect would not occur in sleep-restricted astronauts in space. Therefore, it remains possible that sleep restriction in space could somehow decrease orthostatic tolerance. However, because we saw the opposite trend in this ground-based study, we consider this possibility unlikely.

In conclusion, sleep restriction of 4 h/night for 4 consecutive nights did not affect human orthostatic tolerance, arterial-cardiac baroreflex function, or heart rate variability. Subjects exhibited increased systolic arterial blood pressures and decreased heart rates during simulated orthostatic stress after sleep restriction. This improved maintenance of arterial pressures during LBNP appears to be linked to an increase in peripheral vasoconstriction rather than cardiac output. Thus sleep restriction does produce subtle changes in cardiovascular responses to simulated orthostasis; however, contrary to our hypothesis, these changes do not compromise orthostatic tolerance. Therefore, we conclude that sleep restriction does not decrease orthostatic tolerance, and we offer the possibility that sleep restriction does not contribute to postspaceflight orthostatic intolerance.