Impact of environmental stressors on tolerance to hemorrhage in humans

Abstract

Hemorrhage is a leading cause of death in military and civilian settings, and ~85% of potentially survivable battlefield deaths are hemorrhage-related. Soldiers and civilians are exposed to a number of environmental and physiological conditions that have the potential to alter tolerance to a hemorrhagic insult. The objective of this review is to summarize the known impact of commonly encountered environmental and physiological conditions on tolerance to hemorrhagic insult, primarily in humans. The majority of the studies used lower body negative pressure (LBNP) to simulate a hemorrhagic insult, although some studies employed incremental blood withdrawal. This review addresses, first, the use of LBNP as a model of hemorrhage-induced central hypovolemia and, then, the effects of the following conditions on tolerance to LBNP: passive and exercise-induced heat stress with and without hypohydration/dehydration, exposure to hypothermia, and exposure to altitude/hypoxia. An understanding of the effects of these environmental and physiological conditions on responses to a hemorrhagic challenge, including tolerance, can enable development and implementation of targeted strategies and interventions to reduce the impact of such conditions on tolerance to a hemorrhagic insult and, ultimately, improve survival from blood loss injuries.

INTRODUCTION

Soldiers are frequently required to perform various tasks under adverse conditions, including environmental extremes. Hemorrhage and associated cardiovascular collapse represent a primary cause of battlefield and civilian trauma deaths (35, 57, 113), and the environmental condition to which the soldier is exposed at the time of the hemorrhagic insult may affect the ability to tolerate the injury. An understanding of the effects of the environmental exposure on physiological responses to a hemorrhagic injury is expected to improve detection and subsequent treatment of the hemorrhaging soldier.

Central hypovolemia accompanying hemorrhage occurs secondary to a reduction in venous return to the heart. This leads to a reduction in cardiac output, followed by reduced perfusion pressure and accompanying tissue hypoperfusion and hypoxia. Inadequate perfusion leads to organ (including brain) dysfunction, culminating in multiorgan failure and, ultimately, death. Acute compensatory mechanisms, including elevated sympathetic activity, are engaged to maintain perfusion pressure and, thus, blood flow to vital organs, resulting in increases in cardiac contractility, heart rate, and systemic vasoconstriction. If the environmental condition and the associated physiological state of a human in that environmental condition alter the capacity of these compensatory mechanisms to appropriately respond (either positively or negatively), tolerance to that hemorrhagic insult may be enhanced or impaired, respectively. The objective of this review is to present a current understanding of the impact of the environmental conditions often encountered by the soldier on tolerance to a hemorrhagic insult. Where data are available, physiological responses associated with combined hemorrhagic and environmental stressors are presented. After a methodological discussion of the model used to simulate a hemorrhagic insult in humans (and nonhuman primates), the consequences of passive heat stress, elevated metabolic heat generation, hydration status, cold exposure, and simulated altitude exposure on tolerance to hemorrhagic insult are presented.

LOWER BODY NEGATIVE PRESSURE AS A MODEL OF HEMORRHAGE-INDUCED CENTRAL HYPOVOLEMIA

The concept of lower body negative pressure (LBNP) as a technique to induce central hypovolemia was first introduced into clinical medicine in 1833 by the French physician Victor-Theodore Junod (26). Using a boot-type device, Junod applied negative pressure to one leg and promoted this technique as a method to induce a loss of consciousness for use in surgery (26). Similarly, in the mid-1900s, a number of investigators reported application of LBNP (to lower limbs and torso) to induce hypotensive anesthesia during surgery, subsequently reducing the amount of anesthetic and blood transfusions required and allowing for improved control of arterial pressure (59, 96, 114). In the 1960s, LBNP became a popular method to investigate cardiovascular reflex function in response to shifts in blood volume simulating gravitational stress (i.e., orthostasis) and blood loss (10, 126, 144).

Cooke et al. (20) provided one of the first comprehensive reviews comparing the physiological responses of actual hemorrhage in animals with simulated hemorrhage via LBNP in humans. Although there are key differences between these two stressors, including the loss of blood volume versus sequestration of blood volume in the lower body capacitance vessels and the lack of pain and trauma with LBNP, many common physiological responses between these two stressors were identified. LBNP and hemorrhage reduce venous return and preload and, subsequently, stroke volume and cardiac output (20, 144). During the initial stages of hypovolemia, sympathoexcitation results in compensatory increases in heart rate and systemic vascular resistance that limit the reduction in arterial pressure (20, 144). Hypotension and vital organ hypoperfusion eventually occur at a threshold of volume loss that is highly subject dependent and often (but not always) characterized by sympathetic withdrawal and systemic vasodilation (17, 19, 20). The rate and magnitude of blood loss or LBNP decompression have an impact on the maintenance of arterial pressure, with large and rapid changes resulting in early hypotension and cardiovascular decompensation (20).

One of the first direct comparisons of LBNP with actual blood loss in humans was conducted in 1991 by Rea et al. (95), who removed a total of 450 ml of blood in two equal aliquots and compared the basic hemodynamic and muscle sympathetic nerve activity (MSNA) responses to three mild stages of LBNP (5, 10, and 15 mmHg). LBNP and blood loss decreased central venous pressure (CVP) and increased heart rate and MSNA; the MSNA and CVP responses to 10-mmHg LBNP equated to blood loss of 450 ml (95). In a summary of the available literature, Cooke et al. aligned blood loss from hemorrhage to specific LBNP ranges: 10- to 20-mmHg LBNP equated to 400–550 ml of blood loss (10% of total volume), 20- to 40-mmHg LBNP equated to 550–1,000 ml of blood loss (10–20% of total volume), and ≥40-mmHg LBNP equated to >1,000 ml of blood loss (>20% of total volume) (20).

Most recently, two groups extensively compared the physiological responses of LBNP with actual blood loss within the same subjects in a nonhuman primate model (53) and in humans (60, 97, 132, 133). In the baboon model, Hinojosa-Laborde et al. withdrew 25% of estimated total blood volume in four equal aliquots, equating to 6.25%, 12.5%, 18.75%, and 25%, and then subjected the same animals to four stages of LBNP matched to the pulse pressure and CVP responses elicited by the hemorrhage protocol; the two protocols were conducted on separate days (53). Based on these measures of central blood volume, the four stages of hemorrhage equated to ~20-, 40-, 50-, and 70-mmHg LBNP; 70-mmHg LBNP was equivalent to blood loss of ~18 ml/kg or 1,200–1,300 ml of estimated total blood volume in a 70-kg human (53). As expected, there were differential effects between blood withdrawal and LBNP on hematocrit and hemoglobin, both of which decreased with blood loss due to removal of red blood cells coupled with autoresuscitation-induced hemodilution, while both increased with LBNP due to extravasation of plasma volume and subsequent hemoconcentration (53). The reduction in oxygen-carrying capacity also led to a decrease in venous oxygen saturation with blood removal but no change with LBNP.

In the equivalent human studies, multiple comparisons between LBNP and hemorrhage, including central hemodynamic and hormonal responses (60), cerebral blood flow regulation (97), and responses of coagulation factors (132) and white blood cells (133), have been reported. These data were collected from subjects who were exposed to three progressively increasing stages of LBNP (15, 30, and 45 mmHg) for 5 min each, as well as removal of 1,000 ml of blood in three equal aliquots. These protocols were randomized and separated by a 45- to 75-min recovery period. Overall, the findings of these studies indicate a similar trajectory in most hemodynamic and cerebral blood flow responses (e.g., heart rate, CVP, mean arterial pressure, stroke volume, total peripheral resistance, middle cerebral artery velocity, and cerebrovascular conductance) between LBNP and hemorrhage, but the magnitude of these responses was greater with each step of LBNP compared with hemorrhage. For example, the minimum CVP with LBNP was −0.2 ± 0.6 mmHg versus 1.8 ± 0.8 mmHg with blood loss (132). Similarly, circulating vasoactive hormones, including catecholamines and vasopressin, increased under both conditions, but the magnitude of this increase was greater with LBNP than blood loss (60). An increase in some white blood cell (leukocyte, neutrophil, lymphocyte) concentrations followed a similar pattern (133). In regard to coagulation markers, time to coagulation was accelerated under both conditions, as evidenced by decreased clotting times assessed by thromboelastography (132).

Most commonly, LBNP profiles consist of discrete, progressively decreasing steps in pressure ranging from 2 min (72) to 12 min (16). Less common is the continuous, progressive application of LBNP at variable decompression rates (e.g., 3 mmHg/min), which may more accurately represent continuous-bleeding injuries (4, 18, 61, 63). Both pressure profile categories can be used to elicit presyncope to assess tolerance to progressive central hypovolemia and have demonstrated similar reproducibility in terms of tolerance time and hemodynamic responses (14, 56, 63, 70, 73).

In general, these findings support the continued use of LBNP as a method to simulate preshock hemorrhage in healthy conscious human subjects, particularly in the assessment of environmental stressors (e.g., heat, cold, and altitude) that often accompany hemorrhagic injuries. As previously discussed (26, 126), use of LBNP to simulate hemorrhage in human subjects has many benefits, including the following: 1) it can easily control progressive central hypovolemia at variable rates of decompression, 2) it can progress to maximal tolerance/presyncope in all subjects, 3) it can rapidly reverse the hypovolemic stimulus to facilitate recovery from presyncope, 4) it has no confounding effect of gravity, muscle pump activation, or vestibular activation, as occurs with standing or head-up-tilt maneuvers, and 5) it is a relatively simple, reproducible, noninvasive, and safe technique.

PASSIVE HEAT STRESS AND ITS EFFECTS ON TOLERANCE TO A HEMORRHAGIC INSULT

Soldiers are often exposed to elevated environmental temperatures (1, 11, 12, 131). The physiological consequences of such exposure were outlined in a field study by Buller et al. (11), who instrumented soldiers performing typical mission duties while stationed in Iraq during July/August 2008, when air temperatures were 39–47°C. Soldiers wore standard military uniforms and gear (inclusive of body armor) throughout all assessments. During mounted and unmounted patrols, which were up to 3 h in duration, skin temperatures >38°C and intestinal temperatures approaching 39°C were reported. The impact of such conditions (environmental and physiological) on the ability of humans to tolerate a hemorrhagic insult has been studied by a number of investigators, and their findings are summarized here.

In a number of combat roles, such as sniper or turret gunner, a soldier may be exposed to elevated environmental temperatures for prolonged periods of time, resulting in appreciable increases in skin and core body temperatures. In the laboratory setting, these conditions can be replicated by exposure to similar environmental conditions in a climate chamber or to similar physiological conditions by water-perfused suits. In the case of the water-perfused suit, warm water is perfused through the suit sufficient to increase skin and core temperatures to the desired levels. In both cases, physiological responses to added stressors (such as a simulated hemorrhagic insult) are subsequently assessed. Early work by Horvath and Botelho (55) and later work by Lind et al. (74) clearly showed that tolerance to upright tilt, which, like hemorrhage, reduces central blood volume, was attenuated upon exposure to the applied hyperthermic conditions. Schlader et al. (119) reported on ∼200 tests that assessed tolerance to graded LBNP in normothermic and heat-stressed individuals (Fig. 1). As demonstrated by these data, the ability to tolerate central hypovolemia during this simulated hemorrhagic insult is reduced when individuals are heat-stressed.

Fig. 1.

Tolerance to presyncope during lower body negative pressure (LBNP), quantified by the cumulative stress index, is consistently reduced by passive heat stress sufficient to increase internal temperature by ≥1°C. Cumulative stress index is calculated by summing the product of the LBNP stage and the time at each LBNP stage across the trial until presyncope (i.e., 20 mmHg × 3 min + 30 mmHg × 3 min, etc.). [Republished with permission from Schlader et al. (119).]

Cardiovascular Responses to Passive Heat Stress

Although the aforementioned data provide solid evidence that passive heat stress compromises tolerance to a simulated hemorrhagic insult, less clear are the mechanisms responsible for this occurrence. An understanding of these mechanisms could enable implementation of directed countermeasures to improve tolerance to these combined stressors. Rowell and colleagues performed comprehensive studies that evaluated cardiovascular responses to whole body heat stress in the late 1960s and into the 1970s (102, 106–110) and summarized their work in 1986 (104). Particularly noteworthy is the more than doubling of cardiac output in resting heat-stressed individuals. Such an increase in cardiac output is primarily driven by elevations in heart rate, inasmuch as stroke volume is, for the most part, maintained at normothermic levels in supine individuals. Such an increase in cardiac output is requisite to maintain arterial blood pressure, given reductions in systemic vascular resistance that accompany heat stress due to large reductions in cutaneous vascular resistance. Accompanying these responses are elevations in splanchnic and renal vascular resistances, resulting in reductions in blood flow to these tissue beds. By summing the increase in cardiac output and the reduction in blood flows to the splanchnic and renal vascular beds, Rowell (102) proposed that skin blood flow during profound heat stress can increase from 200–500 ml/min to values approaching 8 l/min.

Rowell (104) hypothesized, though was not able to conclusively demonstrate, that passive heat stress reduced central blood volume secondary to cutaneous vasodilation and accompanying increases in skin blood flow. Using radiographic techniques, Crandall et al. (24, 25) confirmed the hypothesis of Rowell et al. by showing clear reductions in various indexes of central blood volume during whole body heat stress alone, i.e., in the absence of an accompanying simulated hemorrhagic challenge (Fig. 2). Interestingly, they also showed that, for a given degree of simulated hemorrhage (achieved by 30-mmHg LBNP), the magnitude of the subsequent reduction in central blood volume appreciated greatly while subjects were heat-stressed compared with normothermic control conditions (24). Consistent with this hypothesis, Nelson et al. (83) showed smaller (by ~25 ml) left ventricular end-diastolic volumes during upright tilt in heat-stressed individuals than in those exposed to the respective normothermic conditions, while Wilson et al. (137) reported a greater reduction in stroke volume for a given LBNP stage in heat-stressed individuals. Greater reductions in central blood volume, cardiac loading, left ventricular end-diastolic volume, and stroke volume have been proposed to be a primary mechanism by which heat stress compromises tolerance to a simulated hemorrhagic insult (24, 25, 83, 137).

Fig. 2.

Relative reduction in central blood volume in heat-stressed humans, as well as during a non-heat-stress time control trial, in the absence of a simulated hemorrhagic challenge. Note the large reduction in these indexes of central blood volume due to heat stress. Data demonstrate a compromised central blood volume by passive heat stress before an ensuing hemorrhagic insult. Values are means ± SE. [Republished with permission from Crandall et al. (25).]

If the heat stress results in appreciable increases in core body temperature (e.g., greater than ~1.3°C above “normothermia”), cerebral blood flow begins to be compromised (2, 3, 9, 38, 40, 76–78, 81, 85, 87, 118, 129, 139, 140). One study reported a subject with a 33% reduction in cerebral blood velocity when assessed by transcranial Doppler ultrasonography of the middle cerebral artery (9), although, more commonly, the reductions are on the order of 5–20% (3, 9, 40, 85, 88, 129). The primary mechanism responsible for this decrease is reduction in arterial Pco₂ associated with heat-induced hyperventilation (2, 3, 9, 40, 81, 85, 129), although other mechanisms not directly related to arterial hypocapnia (e.g., reductions in blood pressure and, perhaps, neurally mediated vasoconstriction of the cerebral vasculature) have also been implicated (9, 40, 101, 129).

Rowell proposed that the physiological state of a heat-stressed individual should be viewed as a “hyperadrenergic state” (103). This hypothesis was later confirmed by direct recordings of postganglionic MSNA, which increases manyfold during a heat stress (29–32, 64, 75, 86), as well as assessments of the effect of heat stress on left ventricular systolic function (8, 42, 79, 83, 84, 127). Such a hyperadrenergic state induced by heat stress alone (i.e., before the hemorrhagic insult) will reduce the reserve by which sympathetically regulated variables, such as heart rate, cardiac systolic function, and splanchnic/renal vascular resistances, could further increase to assist in the maintenance of cardiac output, systemic vascular resistance, and, thus, arterial pressure during a subsequent hemorrhagic insult. Such responses, when accompanied by heat stress-induced reductions in central blood volume (24, 25), cerebral blood flow (2, 3, 9, 38, 40, 76, 85, 87, 129), and vasoconstrictor responsiveness of the cutaneous vasculature (23, 90, 120, 121), likely contribute to the profound reductions in tolerance to a hemorrhagic insult observed in passively heat-stressed individuals.

Heat Stress, Central Blood Volume, and Tolerance to a Hemorrhagic Insult

The importance of the central hypovolemia that accompanies heat stress in reducing tolerance to a hemorrhagic insult was investigated by Keller et al. (66) and Schlader et al. (118). They exposed subjects to a combination of heat stress and simulated hemorrhage (via LBNP) under “control” conditions and following rapid elevation of central vascular pressures (and, presumably, central blood volume) to pre-heat-stress values via a rapid intravenous infusion of fluid (dextran or hetastarch with saline). In the volume-restored condition, the subjects’ tolerance to the simulated hemorrhagic challenge was completely normalized relative to normothermic subjects (66). Consistent with this observation, the reduction in cerebral blood velocity during the simulated hemorrhagic challenge was attenuated in the volume-loaded state (118). These findings strongly suggest that a primary mechanism of the heat-induced reduction in hemorrhagic tolerance is the reduction in central blood volume that occurs secondary to increases in skin blood flow. Based on those observations, any intervention that would facilitate cutaneous vasoconstriction and, thereby, transfer blood from the skin to the central circulation may be beneficial in improving tolerance to a hemorrhagic insult in a heat-stressed individual. Skin cooling is one such approach that causes cutaneous vasoconstriction with a parallel increase in tolerance to a simulated hemorrhagic challenge (28, 33, 140). Despite this beneficial effect, however, the US Army currently warms hemorrhaging patients, regardless of their core body temperature (15).

Sweating is the primary thermoeffector response to elevations in core body temperature. Depending on the extent of fluid loss from sweating (relative to fluid intake), it is possible that dehydration accompanying a passive heat stress may contribute to compromised tolerance to a hemorrhagic insult. This hypothesis was evaluated by Lucas et al. (80) in subjects whose tolerance to a simulated hemorrhagic insult was assessed following two passive heat-stress exposures: 1) one in which fluid loss due to heat-induced sweating was progressively and precisely offset via continuous intravenous administration of warmed fluid (at a rate of 14 ± 4 ml/min) and 2) one in which such fluid loss was not replenished. Plasma volume was reduced by ~7% in the heat-stress trial without fluid replacement, and this plasma volume reduction was prevented by continuous administration of fluid. Under the fluid replacement condition, tolerance to the simulated hemorrhagic insult improved by ~20% relative to the trial without fluid replacement. Interestingly, even with continuous fluid replacement, tolerance to this hemorrhagic insult remained well below that in a trial in which subjects were normothermic throughout (Fig. 3). Thus, adequate fluid replacement will improve tolerance to a hemorrhagic insult in a passively heat-stressed individual, although tolerance will remain well below that under normothermic conditions.

Fig. 3.

Tolerance to lower body negative pressure (LBNP) challenges during 2 heat-stress trials: 1) without fluid replacement (Dehydration) and 2) with continuous intravenous infusion of body-temperature lactated Ringer solution sufficient to replace sweat loss during the heat stress (IV Infusion). LBNP tolerance during a normothermic time control (Time Controlled) trial was also obtained. Data show a modest contribution of fluid loss due to sweating to reductions in tolerance to a simulated hemorrhagic challenge in heat-stressed individuals. †P < 0.05 vs. normothermic Time Controlled; ¥P < 0.05 vs. Dehydration. See Fig. 1 legend for explanation of the cumulative stress index. Values are means (SD), as well as individual responses. [Republished with permission from Lucas et al. (80).]

Heat Stress, Skin Blood Flow, and Tolerance to a Hemorrhagic Insult

As discussed above, during heat stress the skin vascular bed exhibits a profound increase in conductance. As such, this vascular bed is a potential target to reduce vascular conductance during hemorrhage to maintain arterial blood pressure when cardiac output is reduced due to that insult. However, for reasons that are not entirely clear, such a reflex response is either not elicited, or the extent of cutaneous vasoconstriction is minimal at best. Crandall et al. (23) demonstrated that, during profound reductions in arterial blood pressure associated with a simulated hemorrhagic insult in heat-stressed humans, cutaneous vascular conductance decreased by only ~15% relative to the peak cutaneous vascular conductance during heat stress alone. Moreover, if the site of assessment was locally heated to 38°C, which was the same skin temperature under the water-perfused suit, there were no reductions in cutaneous vascular conductance during the simulated hemorrhagic insult (90). The finding of small, if any, reductions in cutaneous vascular conductance under such conditions strongly suggests that the cutaneous vasculature does not aid in the development of compensatory vascular resistance during a simulated hemorrhagic challenge in heat-stressed humans (43). Little to no cutaneous vasoconstriction occurs despite heat stress generally not impairing baroreflex function (21, 22, 31, 62, 64, 67, 146, 148, 149). Extensive work has been performed to understand the mechanisms underlying the relative absence of cutaneous vasoconstriction in response to a profound hypotensive challenge under heat-stressed conditions; although the precise mechanisms are not fully elucidated, nitric oxide-dependent factors likely contribute (34, 121, 122, 143).

To identify the independent roles of core body and skin temperatures in heat-induced reductions in tolerance to hemorrhagic insult, Pearson et al. (91) exposed subjects to four unique conditions in which core and skin temperatures were independently controlled. It is interesting to note that both core-only heating (i.e., elevated core temperature without elevated skin temperatures) and skin-only heating (i.e., elevated skin temperatures without elevated core temperature) similarly compromised tolerance to a simulated hemorrhagic insult relative to “normothermic” core and skin temperatures. However, the lowest tolerance to the simulated hemorrhagic insult occurred when both core and skin temperatures were elevated. Based on these findings, even large increases in skin temperature, for example, under a soldier’s uniform and associated gear, before parallel increases in core temperature, are sufficient to compromise tolerance to a hemorrhagic insult.

EXERCISE-INDUCED HEAT STRESS AND ITS EFFECT ON TOLERANCE TO A HEMORRHAGIC INSULT

The discussion, thus far, has focused on passive heat stress-induced increases in core body temperature, typically induced via a water-perfused suit in an experimental setting. Although such conditions may mimic those to which a soldier would be exposed when metabolic heat generation is minimal (e.g., stationary in a gun turret or in a sniper position), responses to a hemorrhagic insult may differ if the heat stress is primarily metabolic in origin (i.e., from physical activity/exercise) or combined with hot environmental conditions. Indeed, exercise exacerbates the hypotension response to a subsequent orthostatic challenge (47, 49, 100), and thus one would expect that exercise itself would compromise tolerance to a subsequent hemorrhagic insult. Furthermore, if that exercise was performed under conditions of elevated environmental temperatures, tolerance to a hemorrhagic insult may be even more compromised by these combined stimuli. This question was pursued by Pearson et al. (91), who exposed subjects to 1) passive heat stress and 2) exercise in a hot environment where skin temperature was elevated to the level observed during the passive heat-stress trial. Both of these conditions were followed by a simulated hemorrhagic challenge to maximal tolerance. Just before the hemorrhagic challenge, core body temperatures were similar between the two conditions. Tolerance to the simulated hemorrhagic challenges was not different between the two trials, demonstrating that, when controlling for the elevation in core body and skin temperatures, both passive and active heat stress similarly compromise tolerance.

Similar to a passive heat stress, exercise-induced heat stress causes a reduction in body water and accompanying hypohydration in association with the sweating response. Schlader et al. (117) sought to test the hypothesis that hypohydration accompanying exercise can compromise tolerance to a subsequent hemorrhagic insult. To test this hypothesis, subjects exercised at a workload similar to a military foot patrol in a 40°C and 30% relative humidity environment under the following three conditions: 1) subjects exercised for 90 min and fluid loss was offset by fluid intake (hydrated), 2) subjects exercised for 90 min and fluid intake was prohibited (time-matched hypohydration), and 3) fluid intake was prohibited and the subjects exercised until core body temperature reached the level in the hydrated trial (core temperature-matched hypohydration). Immediately following each of these trials, subjects were exposed to a simulated hemorrhagic insult. When subjects remained hydrated throughout the exercise trial, tolerance to the simulated hemorrhagic insult was greater than in the two hypohydrated trials; there were no differences in tolerance between the two hypohydrated trials (Fig. 4). Thus, in a military setting, soldiers on a foot patrol who remain adequately hydrated during exercise in hyperthermic environmental conditions will be more tolerant to a hemorrhagic insult than those who are not adequately hydrated.

Fig. 4.

Lower body negative pressure (LBNP) tolerance, expressed as the cumulative stress index (see Fig. 1 legend for calculation) following exercise in a hot environment during which 1) water was ingested to offset sweat losses (Hydrated), 2) water was withheld and exercise was terminated upon the same increase in intestinal temperature relative to the hydrated trial (Isothermic Dehydrated), and 3) water was withheld and the exercise duration was the same as that during the hydrated trial (Time Match Dehydrated). Data show that dehydration associated with exercise in the heat will compromise tolerance to a simulated hemorrhagic insult. Values are means (SD). P values are reported for paired comparisons. [Republished with permission from Schlader et al. (117).]

Heat Acclimation and Tolerance to a Hemorrhagic Insult

Heat acclimation is generally achieved by performing exercise in hot environmental temperatures daily for ≥5 days, although some degree of heat acclimation may be achieved by such exposure without exercise (44, 45, 82, 89, 116). The benefits of heat acclimation on thermal tolerance, leading to reduced thermal strain, morbidity, and mortality, as well as cross tolerance, are well known (54, 116). An increase in plasma/blood volume following heat acclimation is proposed to improve orthostatic tolerance (and, thus, may improve hemorrhagic tolerance), perhaps secondary to preservation of stroke volume. However, experimental findings confirming this hypothesis are mixed: a number of studies suggest that heat acclimation improves orthostatic tolerance when the assessment (e.g., upright tilt) is performed in heated conditions (5, 123, 124), while another study did not observe this effect (145). A significant limitation of the studies showing improvements in tolerance after heat acclimation is that a measure of tolerance was not obtained in every subject, given that many subjects did not succumb to upright tilt/standing during pre- and postacclimation trials (5, 123, 124). In one study, where most of the subjects were assessed to tolerance via upright tilting, heat acclimation had no effect on that tolerance (145). Finally, when orthostatic and +G_z acceleration tolerance were assessed in normothermic subjects, +G_z tolerance was unaffected by 12 days of heat acclimation but orthostatic tolerance (via upright tilting) was improved in the men only after heat acclimation (46). Since heat acclimation is recommended before a soldier encounters hyperthermic environmental conditions (130), further knowledge regarding the potential benefits of that acclimation regimen in improving tolerance to a hemorrhagic insult would be particularly insightful.

HYPOTHERMIC EXPOSURE AND TOLERANCE TO A HEMORRHAGIC INSULT

Particularly in mountainous regions, soldiers often encounter hypothermic environmental conditions. In the laboratory setting, hemodynamic responses to whole body skin surface cooling are often mixed, likely due to varying intensities of the cooling stimuli, the duration of the cooling stimuli, and whether the cooling stimulus was sufficient to cause shivering. One relatively consistent observation accompanying whole body cooling is an elevation of central vascular and arterial blood pressures (7, 27, 28, 33, 37, 52, 65, 92, 93, 138, 141, 142). This response primarily occurs through a combination of elevations in cardiac output and peripheral vascular resistance (7, 28, 50, 65, 92, 125, 141), although some studies report a reduction in peripheral vascular resistance during prolonged cooling, despite sustained elevations in blood pressure (92). Despite this latter observation, plasma norepinephrine concentrations increase during a cooling stimulus (33, 136), while MSNA typically does not change (27, 37, 65). It is noteworthy that moderate skin surface cooling also increases cerebral blood velocity, likely secondary to the accompanying elevations in arterial blood pressure (33, 65, 147).

A number of studies have investigated the effects of whole body cooling on autonomic responses to a simulated hemorrhagic challenge (28, 33, 65, 93, 94, 140), although most of these studies did not assess tolerance to this stress. The first to explore the latter question was Durand et al. (33), who assessed tolerance to a progressive LBNP protocol to presyncope in otherwise normothermic subjects. The application of skin surface cooling, via water-perfused suits, before and throughout the simulated hemorrhagic challenge increased tolerance by ~34% relative to the normothermic control condition. Skin surface cooling before and throughout the simulated hemorrhagic challenge enhanced blood pressure and cerebral blood velocity during LBNP, while the increase in calf volume in response to LBNP was attenuated with the cooling stimulus (Fig. 5). Particularly interesting was the approximate doubling of plasma norepinephrine concentrations at 40-mmHg LBNP, as well as at the end of the LBNP challenge, when performed with skin surface cooling (33).

Fig. 5.

Changes in calf volume, mean arterial blood pressure, and cerebral blood velocity before and during progressive lower body negative pressure (LBNP) to presyncope while subjects were normothermic and during skin surface cooling (SSC). Numbers in parentheses depict the number of subjects analyzed for each of the LBNP stages; data were included only if responses were obtained for the specified LBNP stage during normothermic and SSC trials. †P < 0.05 vs. normothermia for specified LBNP stage; *P < 0.05, baseline (Base) vs. SSC (i.e., before LBNP). From Durand et al. (33).

Cui et al. (28) sought to identify the mechanisms by which skin cooling improves tolerance to a simulated hemorrhagic challenge. Although the skin surface cooling stimuli did not alter baroreflex responsiveness (27), it did attenuate the reductions in CVP and stroke volume throughout the ensuing simulated hemorrhagic challenge (28). They concluded that this upward shift in CVP during skin cooling preserved cardiac preload, resulting in less reduction in stroke volume and arterial blood pressure for a given magnitude of simulated hemorrhage. Similarly, a subsequent study by Wilson et al. (142) demonstrated elevations in pulmonary capillary wedge pressure (an index of left ventricular filling pressure) and pulmonary arterial pressure when skin surface cooling accompanied the simulated hemorrhagic challenge, demonstrating an increase in cardiac preload.

A gap in the literature is an investigation into the effects of environmental cooling, with and without a metabolic heat load, on the capacity for individuals to tolerate a hemorrhagic insult. Depending on the extent of environmental cooling, coupled with the insulating capabilities of clothing, skin temperatures during such exposure may be appreciably higher than those in the aforementioned studies of partially nude subjects cooled in a water-perfused suit. Thus it would be reasonable to hypothesize that if skin temperatures were higher during physical activity, such as a foot patrol, any benefits of the environmental cooling stimulus in preserving tolerance to a hemorrhagic insult would be minimized. Another gap in our knowledge in these areas is the extent to which core body temperature cooling, with and without accompanying skin surface cooling, may influence tolerance to a hemorrhagic insult.

ALTITUDE/HYPOXIA EXPOSURE AND TOLERANCE TO A HEMORRHAGIC INSULT

Mountain or alpine warfare has been fairly common throughout history. Several important conflicts, such as those during World War I in the Italian Alps, have taken place at >3,000-m altitude. Also, many areas in South and Central Asia, such as the mountainous areas of Afghanistan and Pakistan, have been sites of more recent conflicts. Presumably, many special operations missions have also been conducted in mountainous regions in other areas throughout the world. Mountain warfare poses several physiological challenges to the soldier, including cold stress and the hypobaric environment. Hypobaria causes a direct reduction in blood oxygen content and can subsequently attenuate oxygen delivery if the reduction in blood flow and/or arterial oxygen content is severe enough. For instance, oxygen content is ~17 ml/dl (a ~15% reduction from sea level) at an altitude of 3,000 m. The effect of the reduced oxygen content on oxygen delivery to the tissues is further exacerbated under conditions of reduced blood flow, such as a hemorrhagic insult. The compensatory responses to low arterial oxygen content at rest include increases in sympathetic activity, cardiac output, and mean arterial pressure and decreases in systemic vascular resistance (135). These responses are elicited to augment perfusion pressure and blood flow to defend against tissue hypoxia, which is the main concern during blood loss in hypobaric environments.

Soldiers injured at altitude are likely predisposed to an increased risk of cardiovascular decompensation to a hemorrhagic injury due to the systemic effects of hypoxemia. Heistad and Wheeler (51) investigated the combined effects of LBNP and hypoxia on blood pressure and forearm vascular resistance. Subjects were exposed to 40-mmHg LBNP for 90 s while breathing 10% oxygen, 12% oxygen, or room air (~21% oxygen). During 40-mmHg LBNP and 12% oxygen breathing, mean arterial pressure decreased and the increase in forearm vascular resistance was attenuated relative to the room air condition. The reductions in mean arterial pressure and the attenuated rise in forearm vascular resistance during LBNP were exacerbated in subjects breathing 10% oxygen versus 12% oxygen. The exaggerated responses to the brief LBNP challenge during 10% oxygen breathing, compared with the 12% oxygen and room air conditions, suggest that exposure to higher elevations might exacerbate the reduced ability to tolerate a hemorrhagic insult compared with sea-level or lower-altitude conditions.

Although they have not been validated as a technique to study the physiological effects of blood loss, head-up-tilt challenges have been used by several investigators to evoke central hypovolemia during hypoxic conditions. Rickards and Newman found that 17.1% oxygen and 14.2% oxygen breathing, to simulate 1,680- and 3,350-m altitudes, respectively, did not influence the ability to complete 2 min of a 75° head-up-tilt challenge (98). Despite the short duration of the head-up-tilt challenge, the increase in heart rate was attenuated during 14.2% oxygen breathing relative to a normoxic trial, although mean arterial pressure was well maintained. When compared with control conditions, an increase in the incidence of subjects unable to complete 10- to 20-min head-up-tilt challenges at 3,660-m (6), 3,700-m, and 4,300-m (112) simulated altitudes in a hypobaric chamber or while breathing normobaric hypoxic gas with 12% oxygen (48) has been reported. Subjects who were unable to complete the head-up-tilt challenges in normobaric hypoxic conditions exhibited an attenuated rise in forearm vascular resistance (48).

Rowell and Blackmon (105) also reported blunted increases in forearm vascular resistance, as well as greater reductions in mean arterial pressure, in subjects exposed to 10 min of 30- to 40-mmHg LBNP while breathing 10% oxygen relative to a room air condition. The extended duration of LBNP (10 min at −30 to −40 mmHg) induced greater reductions in mean arterial pressure and higher heart rates relative to 90 s of 40-mmHg LBNP in subjects breathing 10% oxygen (51). This suggests that the duration of these combined conditions (i.e., hemorrhagic insult and hypoxic duration) likely contributes to cardiovascular collapse in the bleeding soldier at altitude. Additionally, four of the eight subjects tested by Rowell and Blackmon exhibited signs of cardiovascular decompensation that were accompanied by a ∼10-fold increase in circulating epinephrine at the termination of LBNP and hypoxia versus the normoxic trial, which did not augment circulating epinephrine. These findings indicate that systemic β-adrenergic-stimulated vasodilation may contribute to a reduced ability to maintain blood pressure during a simulated hemorrhagic insult in hypoxic conditions in some subjects. In this context, it has been estimated that ∼50% of resting hypoxic vasodilation is due to β-adrenergic stimulation (135). Therefore, subjects who exhibit large increases in circulating epinephrine are most likely at a greater risk of hypotension during a hypoxic hemorrhagic challenge due to impaired vasoconstriction. Rowell and Seals (111) furthered the work of Rowell and Blackmon by inducing graded reductions in central blood volume (5-min stages at 5-, 10-, 15-, 20-, and 25-mmHg LBNP) while subjects breathed 10–12% oxygen. Despite a relatively stable mean arterial pressure throughout LBNP and hypoxia, the rise in forearm vascular resistance was attenuated in hypoxia relative to the normoxic trial. Furthermore, heart rate and MSNA were greater throughout LBNP when combined with hypoxia versus normoxia. Similar to the earlier work of Rowell and Blackmon (105), large increases in circulating epinephrine were observed concomitantly with decreases in mean arterial pressure, heart rate, and forearm vascular resistance in three of eight subjects. Teague and Hordinsky used graded LBNP (5-min stages at 15-, 30-, and 45-mmHg LBNP) in the seated posture to evoke central hypovolemia in subjects at sea level and at 3,810-m simulated altitude in a hypobaric chamber (128). All subjects were able to complete the LBNP protocol at sea level. However, 30% could not complete the LBNP protocol at altitude due to cardiovascular collapse (i.e., blood pressure <90/60 mmHg, heart rate <60 beats/min, and/or a fall in diastolic middle cerebral artery blood velocity) or symptoms such as dizziness, nausea, and/or light-headedness. The subjects who were not tolerant to LBNP at altitude demonstrated lower systemic vascular resistance and attenuated increases in heart rate, which resulted in greater reductions in mean arterial pressure than in subjects who were able to tolerate the LBNP protocol at altitude. Collectively, the reduction in oxygen content at altitude appears to impair tolerance to a simulated hemorrhagic insult by blunting the rise in vascular resistance. Furthermore, some subjects might be more likely to experience an earlier onset of cardiovascular collapse during a hemorrhagic insult at altitude due to a surge of circulating epinephrine and the resulting systemic β-adrenergic-stimulated vasodilation that contributes to the decrease in blood pressure in these subjects.

In addition to the relative inability to sufficiently increase systemic vascular resistance during LBNP with hypoxia, a reduction in cerebral blood flow might also play a role in determining tolerance to a hemorrhagic insult at altitude. Acute hypoxia stimulates hyperventilation and a subsequent reduction in arterial Pco₂. In turn, reductions in arterial Pco₂ lower cerebral blood flow (58, 69) and could attenuate tolerance to a hemorrhagic insult. Teague and Hordinsky also examined the middle cerebral artery blood velocity responses in individuals subjected to graded LBNP at 3,810-m simulated altitude in a hypobaric chamber (128). They reported greater reductions in middle cerebral artery blood velocity in subjects who did not complete the LBNP challenge at altitude than in those who were able to tolerate the LBNP challenge at altitude. van Helmond et al. (134) also examined the hemodynamic and cerebrovascular responses to graded LBNP (5-min stages at 15-, 30-, and 45-mmHg LBNP) during normoxia and acute hypoxia with arterial oxygen saturation clamped at 85%. During LBNP combined with hypoxic conditions, mean arterial pressure and cardiac output were attenuated, while oscillations in mean arterial pressure and middle cerebral artery blood velocity (i.e., an assessment of the magnitude of hemodynamic fluctuations within a defined frequency range) were increased compared with normoxia. However, arterial Pco₂, mean middle cerebral artery blood velocity, and cerebral autoregulation were not different between hypoxic and normoxic conditions. The greater increase in low-frequency (0.04- to 0.15-Hz) mean arterial pressure oscillations during LBNP in hypoxia than normoxia likely reflects an exacerbated elevation in sympathetic nerve activity during hypoxia, which is consistent with the previous work by Rowell and Seals (111). These oscillations in mean arterial pressure are likely then passively transferred to the cerebral vasculature, accounting for the similar increase in middle cerebral artery blood velocity low-frequency oscillations. It has been speculated that these oscillations may contribute to the maintenance of cerebral tissue perfusion and oxygenation during central hypovolemia (99), although this has yet to be directly assessed. Although four of nine subjects did not complete the LBNP-in-normoxia trial and six subjects did not complete the LBNP-in-hypoxia trial, van Helmond and colleagues (134) did not compare the physiological responses to LBNP in hypoxia between tolerant and intolerant subjects.

Although the available data indicate that LBNP tolerance is reduced during hypoxia, none of the aforementioned studies were designed to determine the magnitude by which hypoxia reduces maximal tolerance to a hemorrhagic challenge in all subjects. So while the physiological responses to a simulated hemorrhagic insult during hypoxia appear to predispose a soldier to cardiovascular collapse at an attenuated central hypovolemic state relative to sea level, this issue has not been specifically addressed. Additional studies are needed to assess maximal tolerance to a simulated hemorrhagic insult at sea level and altitude (or hypoxia). Along these lines, hypoxia was introduced a short time (<60 min) before commencement of the LBNP protocols in the reviewed studies. These short hypoxic exposures likely do not reflect the hypobaric conditions to which a soldier would be exposed during prolonged missions. Several physiological adaptations to altitude, such as a rightward shift in the oxygen-hemoglobin dissociation curve and increases in hemoglobin and hematocrit, improve oxygen-carrying capacity at altitude, which conceivably would improve tolerance to a hemorrhagic insult. However, plasma volume is typically decreased during altitude acclimatization (115), and reductions in plasma volume attenuate LBNP tolerance (117).

To address the potential problem of altitude acclimatization and tolerance to central hypovolemia, Fulco and colleagues exposed eight subjects to 13 min of a 60° head-up-tilt challenge at sea level and after 1, 18, 66, and 114 h at 4,300-m simulated altitude in a hypobaric chamber (41). After 66 and 114 h of acclimatization, resting hemoglobin and hematocrit were greater than sea level values, but calculated blood and plasma volumes were lower than sea level values. Resting systemic vascular resistance was greater than at sea level at 66 and 114 h of acclimatization, and mean arterial pressure was greater than at sea level at 18 and 114 h of acclimatization. However, systemic vascular resistance during the head-up-tilt challenge after 1 h of acclimatization was reduced by 22% but was not different from sea level after 18 or 66 h of acclimatization. Mean arterial pressure during the head-up-tilt challenge was lower than at sea level after 1 h of altitude exposure but was greater than at sea level following >18 h of acclimatization. Because all the subjects were able to tolerate the entire head-up-tilt challenge at each stage of acclimatization, it is unclear if altitude acclimatization improves tolerance to a simulated hemorrhagic insult. Nevertheless, the altitude acclimation-induced changes in resting hemoglobin and hematocrit, as well as the improvements in systemic vascular resistance and mean arterial blood pressure during the head-up-tilt challenge, are likely to improve tolerance to a hemorrhagic insult at altitude.

WHAT REMAINS UNKNOWN AND IMPORTANT AVENUES OF FUTURE RESEARCH

As detailed in this review, the impact of many of the individual environmental stressors that accompany hemorrhagic injuries on physiological responses to LBNP and LBNP tolerance, including passive heat stress, dehydration, elevated metabolic heat generation, cold exposure, and acute hypoxic exposure, has been assessed. Surprisingly, studies assessing the impact of acute and chronic high-altitude exposure on maximal tolerance to hemorrhagic stress are lacking and would be an important contribution to this field. An additional area that has been less explored is the impact of combined environmental stressors, such as cold and altitude (hypoxia), cold and elevated metabolism, and dehydration and hypoxia, on these responses; the possible combination of factors is extensive. Also, in the microgravity literature, physical deconditioning associated with prolonged bed rest and, perhaps, microgravity exposure itself are known to reduce tolerance to upright tilt and/or LBNP (13, 36, 39, 65, 68, 71, 150). Although it is unlikely that hemorrhage will be a problem following spaceflight, the potential negative impact of deconditioning on hemorrhagic tolerance remains to be fully elucidated. Furthermore, the role of many factors that alter baseline physiological status, including sleep deprivation/restriction, negative energy balance, chronic and acute nicotine and caffeine use, acute exercise/intense physical activity, and acclimatization to heat, cold, and/or altitude, has not been fully addressed. Further work should address the impact of these conditions on key regulatory responses to hemorrhage, including hemodynamic, autonomic, and cerebrovascular regulation, inflammation and oxidative stress markers, and the coagulation pathway. An understanding of the impact of these environmental and behavioral stressors on the physiological responses to hemorrhage will facilitate the development of strategies and interventions that may improve hemorrhage tolerance and, ultimately, increase the likelihood of survival from blood loss injuries on the battlefield.

GRANTS

Much of the authors’ work outlined in this review was supported by the National Heart, Lung, and Blood Institute, the National Institute of General Medical Science, the Department of Defense, and the American Heart Association.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

C.G.C., C. A. R., and B.D.J. interpreted results of experiments; C.G.C., C.A.R., and B.D.J. drafted manuscript; C.G.C., C.A.R., and B.D.J. edited and revised manuscript; C.G.C., C.A.R., and B.D.J. approved final version of manuscript.