OHN G. LAFFEY, M.D., AND BRIAN P. KAVANAGH, M.B.
ARTERIAL carbon dioxide tension represents the balance between the production and elim-ination of carbon dioxide, and in healthy persons, it is maintained within narrow physiologic limits. Hypocapnia, even when marked, is normally well tolerated, often with few apparent effects. Tran-sient induction of hypocapnia can lead to lifesaving physiological changes in patients with severe intra-cranial hypertension or neonatal pulmonary-artery hy-pertension, but hypocapnia of longer duration in crit-ically ill patients may adversely affect outcomes.1,2 Despite concern about adverse effects, the induction of hypocapnia has commonly been recommended for diverse disease states.3-5 Thus, hypocapnia, whether produced deliberately3-5 or accidentally,6,7 remains prevalent in clinical practice (Table 1). In addition, hypocapnia is a common component of many acute illnesses, although its importance is often underesti-mated.8-12 The prevalence of hypocapnia may be ex-acerbated by the belief held by some clinicians that hypocapnia is inherently safer than — or at least preferable to — hypercapnia.
DEVELOPMENT OF ARTERIAL
In its simplest form, the partial pressure of arterial carbon dioxide (PaCO2) reflects the balance be-tween the production and elimination of carbon di-oxide (CO2), as described by the following formula:
PaCO2 is proportional to CO2 production + inspired CO2 .
The volume of inspired carbon dioxide is usually negligible, whereas reduced carbon dioxide production is an unusual, but possible, contributor to hypo-capnia. Therefore, for practical purposes, a low partial pressure of arterial carbon dioxide reflects the rate of elimination of carbon dioxide. Thus, the principal physiologic causes of hypocapnia, including pregnan-cy, are related to hyperventilation (Table 1).13 Of course, hyperventilation can occur with mechanical ventilation, and artificial clearance of carbon dioxide with the use of extracorporeal techniques (e.g., car-diopulmonary bypass, extracorporeal membrane oxy-genation, or devices for the removal of carbon diox-ide) is extraordinarily efficient.6,14,15
One form of hypocapnic alkalosis that is rarely dis-cussed occurs during critical reduction of pulmonary perfusion (for example, during cardiopulmonary re-suscitation). In such cases, there is a dissociation be-tween the condition of central venous blood, with a high partial pressure of arterial carbon dioxide and a low pH, and that of the systemic arterial blood, with a low carbon dioxide tension and an alkalemic pH; this dissociation is due to the combination of low pul-monary perfusion and normal ventilation, and this condition is called pseudorespiratory alkalosis.16
Therapeutic Induction of Hypocapnia
The deliberate induction of hypocapnia for short periods while other definitive treatment measures are being instituted remains a potentially lifesaving ther-apeutic strategy in situations in which intracranial pres-sure17,18 or neonatal pulmonary vascular resistance is critically elevated. There is no evidence to support the therapeutic or prophylactic use of induced hypo-capnia in any other context. However, induced hy-pocapnia has been and may remain a common prac-tice, particularly in patients with brain injury or neonatal respiratory failure, as well as during general anesthesia.
The recognition of the effectiveness of hypocapnic alkalosis in reducing intracranial pressure, together with the identification of elevated intracranial pres-sure as a pathogenic condition, led to the assumption that hypocapnia should be induced when intracranial pressure was elevated (Fig. 1). Therefore, hyperventila-tion20-24 — sometimes resulting in very severe hypo-capnia21-24 — once represented the standard of care for the treatment of patients with head trauma.
Despite expert guidelines recommending against it and evidence of adverse outcomes, deliberate hy-perventilation continues to be widely practiced.25-27 In the United States, 36 percent of board-certified neurosurgeons25 and almost 50 percent of emergency physicians26 routinely use prophylactic hyperventila-tion in patients with severe traumatic brain injury. The suggested indications for its use continue to vary: some suggest using it for suspected,28 established,27 or intractable29 intracranial hypertension, whereas others recommend that it be used only for intracranial hyper-tension that is accompanied by neurologic deterio-ration.30,31 In addition, contemporary textbooks rec-ommend the induction of substantial hypocapnia (partial pressure of arterial carbon dioxide of approx-imately 25 mm Hg) as a preliminary measure after severe head trauma in both adults3 and children.
Other Forms of Coma
Because of its effects on intracranial pressure, hy-perventilation has been advocated for the manage-ment of coma after near-drowning4 or near-hanging, as well as for the management of cerebral edema in patients with diabetic ketoacidosis.32 The latter rec-ommendation is of particular concern in the light of the recent recognition of hypocapnia as a key predictor of the development of cerebral edema in chil-dren with diabetic ketoacidosis.
Neonatal respiratory failure commonly involves pulmonary hypertension, right-to-left shunting, and profound hypoxemia. Hyperventilation to relieve pul-monary hypertension has been advocated for neo-nates with persistent pulmonary hypertension of the newborn19,34 or with congenital diaphragmatic her-nia.35 In addition, in the resuscitation of neonates, hyperventilation could rapidly clear excess carbondioxide resulting from bicarbonate administration and could counteract metabolic acidosis.36 Previous rec-ommendations included prolonged maintenance of a partial pressure of arterial carbon dioxide below 20 to 30 mm Hg in such infants.
Anesthesia and Surgery
Moderate to severe hypocapnia (partial pressure of arterial carbon dioxide, 20 to 25 mm Hg) has, in the past, been widely advocated as an adjunct to gen-eral anesthesia.38 Its proposed advantages include the minimization of spontaneous respiratory effort and a reduced requirement for sedative, analgesic, and mus-cle-relaxant medications.38 The latter advantage may explain the widespread use of intraoperative hyper-ventilation in the 1960s39 as a means of reducing the use of anesthetic medications and thus avoiding fetal depression immediately after cesarean section. The use of hypocapnia during general anesthesia remained common for at least the next two decades.
Accidental Induction of Hypocapnia
Hypocapnia can develop as a result of excessive mechanical ventilation.2,7,35 In addition, cardiopulmo-nary bypass,6 high-frequency modes of ventilation,14 and extracorporeal membrane oxygenation15 have been associated with the development of unanticipat-ed hypocapnia. The common use of these techniques in neonates, coupled with the potential for hypocap-nia-associated intraventricular hemorrhage, suggests that neonates may represent the most vulnerable subgroup of patients (Fig. 2). Because clearance of metabolic acids from the cerebrospinal fluid after he-modialysis takes longer than systemic clearance, hyperventilation may occur, causing hypocapnic alkalo-sis in patients receiving long-term hemodialysis.
Hypocapnia as a Common Component of Disease
Hypocapnia is also an inherent component of several disease states (Table 1) and is a consistent finding in patients with early asthma,11 high-altitude pulmo-nary edema,12 or acute lung injury.41 Hypocapnia has long been recognized as the most common acid– base disturbance in critically ill patients,9 and it is a consistent feature of both septic shock10 and the sys-temic inflammatory response syndrome.8 In fact, hypocapnia is a diagnostic criterion for the latter condition.8 In addition, it is a prominent feature of diabetic ketoacidosis in children and is a key predictor of cer-ebral edema in such children.33
PATHOBIOLOGY OF HYPOCAPNIA
When it is mild, hypocapnia does not have serious effects in healthy persons. Symptoms and signs in-clude paresthesias, palpitations, myalgic cramps, and seizures.42 However, extensive data from a spectrum of physiological systems indicate that hypocapnia has the potential to propagate or initiate pathological processes. As a common aspect of many acute disor-ders, hypocapnia may have a pathogenic role in the development of systemic diseases.
Hypocapnia, Hypocapnic Alkalosis, and Acid–Base Status
Hypocapnic alkalosis is synonymous with respira-tory alkalosis. Acute hypocapnia results in the imme-diate development of alkalosis; at any given moment, the extracellular pH may be predicted on the basis of the Henderson–Hasselbach formula:
The buffering response to acute hypocapnia is bi-phasic. First, hypocapnia in the extracellular fluid re-sults in an immediate decrease in the intracellular-fluid carbon dioxide concentration, resulting in the transfer of chloride ions from the intracellular fluid to extracellular-fluid compartments. This chloride-ion egress, accompanied by a decrease in the con-centrations of bicarbonate ions in extracellular fluid, is called tissue buffering.16 Second, the renal re-sponse (inhibition of renal tubular reabsorption of bicarbonate ions) can begin within minutes and takes effect over a period of hours to days.16 With long-term exposure, in the presence of normal renal function, the bicarbonate-ion level begins to fall, and the pH decreases but does not reach the normal value of 7.4 (i.e., a hydrogen ion concentration of 40 nmol per liter).
Respiratory versus Metabolic Alkalosis
The clinical physiology of acid–base disorders fo-cuses on the conditions in the extracellular-fluid compartment. The carbon dioxide molecule is more lipid-soluble than the hydrogen ion, and therefore, acid–base alterations arising from an altered partial pressure of arterial carbon dioxide (respiratory alka-losis or respiratory acidosis) equilibrate across cell membranes (i.e., between extracellular and intracel-lular fluid) far more rapidly than do primary meta-bolic acid–base changes. Thus, at a given extracellu-lar pH, the cellular effects are more pronounced when the alkalosis has a respiratory basis than when it has a metabolic basis. Nonetheless, most effects of extracellular hypocapnia result from alkalosis rather than from a low partial pressure of arterial carbon dioxide itself, as has been documented with respect to pulmonary,43 cerebral,44 and placental45 perfusion, as well as myocardial effects.46 Finally, an additional physiologically-based approach to the analysis of hy-drogen-ion homeostasis, called the “strong ion difference” and initially described by Stewart, has been reviewed extensively.47 According to this approach, the only factors that determine the pH reflect conserva-tion of mass and electrochemical neutrality. These factors can be reduced to the following three groups: the strong ion difference (the sum of the concentrations of sodium, potassium, calcium, and magnesium minus the concentrations of chloride and lactate), the concentration of weak acids (proteins and phosphates), and the partial pressure of arterial carbon dioxide.
Hypocapnia, Cellular Metabolism, and Oxygenation
At the tissue level, an oxygen imbalance occurs when oxygen demand (which reflects the metabolic rate) outstrips oxygen supply. Hypocapnia may cause or aggravate cellular or tissue ischemia by both de-creasing the cellular oxygen supply and increasing the cellular oxygen demand (Fig. 3). Although hypocapnia induced by hyperventilation may increase alveolar oxygen tension, multiple important pulmonary effects of hypocapnic alkalosis (e.g., bronchoconstriction,48 attenuation of hypoxic pulmonary vasoconstriction,49 and increased intrapulmonary shunting49) result in a net decrease in the partial pressure of arterial oxygen. Because both hypocapnia and alkalosis cause a left-ward shift of the oxyhemoglobin dissociation curve, off-loading of oxygen at the tissue level is restrict-ed.50 In addition, hypocapnia causes systemic arterial vasoconstriction, decreasing the global and regional oxygen supply and compounding the reduction in the delivery of oxygen to tissue.51
Hypocapnia may increase the metabolic demand of tissue through cellular excitation or contraction (Fig. 3). Finally, alkalosis — especially respiratory al-kalosis — inhibits the usual negative feedback by which
a low pH limits the production of endogenous organic acids (such as lactate).52
Dose–Response Relation and Duration of Hypocapnia
Although mild hypocapnia results in few or no se-rious effects, marked hypocapnia may cause serious adverse effects.30,50,53-55 However, data are limited, and extrapolation to all organ systems or disease entities might not be justified. If hypocapnia is prolonged, buffering (by decreasing the level of bicarbonate ions in extracellular fluid) results in a gradual return of the extracellular fluid pH toward normal. In the brain, because local pH determines the degree of cerebral vasoconstriction, such buffering normalizes cerebral blood flow,44 decreasing the effectiveness of the re-duction in intracranial pressure56 and possibly atten-uating the neuronal ischemia. This scenario in the central nervous system is complicated, because the restoration of the partial pressure of arterial carbon dioxide to normal after buffering may result in cere-bral hyperperfusion56,57 that can cause a rebound in-crease in intracranial pressure, aggravate reperfusion injury (Fig. 1), or precipitate hemorrhage (Fig. 2).
HYPOCAPNIA AND THE BRAIN
The mechanisms underlying the adverse neurologic consequences of hypocapnia are similar to those seen in other tissues when there is reperfusion injury and an imbalance between oxygen supply and demand. The control of acid–base homeostasis in the cere-brospinal fluid has been reviewed extensively,58 with special attention to important specific issues in the regulation of the cerebral circulation.
The cranial cavity has a fixed volume, and when the mass of any of its contents increases (as it may, for example, in patients with cerebral edema, hematoma after a head injury, or a brain tumor), a critical eleva-tion of intracranial pressure may occur. This elevation in pressure may result in impaired cerebral perfusion, a risk of brain-stem herniation, and possibly, adverse outcomes from direct pressure on neuronal cells (Fig. 1). In order to reduce intracranial pressure, the volume of the cranial contents must be reduced. Hy-pocapnic alkalosis decreases the cerebral blood volume by means of potent cerebral vasoconstriction, thereby lowering intracranial pressure (Fig. 1).
Mechanisms of Deleterious Central Nervous System Effects
The beneficial effects of hypocapnia on intracrani-al pressure, however, may be outweighed by the ef-fects of a reduced oxygen supply. If the reduction in cerebral blood flow is disproportionately greater than that of the intracranial blood volume,59 cerebral ische-mia can result.60 In experimental cerebral ischemia,
Figure 3. Effects of Hypocapnia on Global Oxygen Supply and Demand.
Hypocapnic alkalosis adversely alters the balance between global oxygen delivery and oxygen consumption. It decreases global and regional oxygen delivery through a combination of decreased systemic oxygen tension, tissue perfusion, and oxy-gen unloading at the tissue level. Conversely, hypocapnic alka-losis may increase the metabolic requirement for oxygen at the cellular level through physiological increases in cell excitation or contraction. It may also directly contribute to the pathogen-esis of acute lung injury and systemic inflammation. These in-terrelated actions of hypocapnic alkalosis may critically com-promise cellular survival and contribute to adverse outcomes.
hypocapnia increases the lactate production associat-ed with ischemia,61 although this may be explained in part by an inhibition of brain phosphofructokinase activity that is unrelated to ischemia.62 In the past, hypocapnia was thought to increase regional perfu-sion to ischemic parts of the brain at the expense of uninjured brain tissue; this phenomenon, termed “inverse steal,” does not actually occur.63 In fact, hy-pocapnia increases cerebral oxygen demand. Hypo-capnia increases neuronal excitability, seizure activity,64 and anaerobic metabolism.61 Finally, the presence of hypocapnia during cardiopulmonary resuscitation may worsen brain injury.65 Hypocapnic potentiation of seizure activity, in addition to increasing oxygen demand, augments production of the cytotoxic ex-citatory amino acids associated with seizures.66 Hy-pocapnia may also induce increases in neuronal dopa-mine,67 which may increase the risk of convulsions.
Deleterious Central Nervous System Effects in Clinical Context
Neonatal Brain Injury
Hypocapnia appears to be particularly injurious to the brain in premature infants (Fig. 2). In preterm
infants who are exposed to severe hypocapnia (a par-tial pressure of arterial carbon dioxide of less than 15
mm Hg [<2 kPa]), even of relatively short duration, considerable long-term neurologic abnormalities may develop68; such abnormalities are associated with many forms of pathologic neonatal brain conditions (Table 2). Neurovascular factors that may predispose the immature brain to such injury include poorly de-veloped vascular supply to vulnerable areas,69 antioxi-dant depletion by excitatory amino acids,70 and the lipopolysaccharide71 and cytokine72 effects that poten-tiate destruction of white matter (Fig. 2).
Data from neonates clearly suggest that severe hy-pocapnia after hyperventilation,68 high-frequency ven-tilation,14 and extracorporeal membrane oxygenation15 contribute to adverse neurologic outcomes. In addi-tion, abrupt termination of hyperventilation results in reactive cerebral hyperemia,57 which may cause in-tracranial hemorrhage in premature neonates.57 Be-cause hypocapnia, induced by accident or design, is common in such neonates, awareness of these asso-ciations is extremely important.
Traumatic Brain Injury
In patients with traumatic brain injury, prophylac-tic hyperventilation is actually associated with worse outcomes,1 which may be explained in part by re-duced cerebral oxygenation.55 Thus, although intra-cranial pressure may decrease transiently, it may do so at the expense of cerebral perfusion.59 In addi-tion, hypocapnia may exacerbate secondary brain in-jury, because increased cerebral vascular reactivity and vasoconstriction can result in decreases in re-gional cerebral blood flow.60 Therefore, hypocapnia may result in a disproportionate (regional) decrease in cerebral blood flow, without a further decrease in intracranial pressure.18 Because of these possibilities, a panel of experts has recommended against the pro-phylactic use of hyperventilation.30
Hyperventilation has classically been advocated as a therapy for patients with acute stroke, intended to reduce intracranial pressure (Fig. 1), to induce in-verse steal in ischemic areas of the brain, and to cor-rect acidosis in the zones around ischemic tissue. Despite the theoretical physiological benefits, as de-scribed earlier, positive outcomes have not been re-alized; rather, patients do poorly.73
Postoperative Psychomotor Dysfunction
Cognitive impairment after general anesthesia is a cause for concern, because the growing trend to-ward ambulatory (same-day) anesthesia and surgery (and away from overnight stays in hospitals after sur-gery) results in the discharge of patients at an earlier
stage of their recovery. Postoperative cognitive dys-function is of particular concern in the elderly, who are more susceptible to it and more vulnerable to its consequences. Acute hypocapnia is common during general anesthesia, and otherwise healthy patients who are subjected to hypocapnia during general anesthe-sia have been found to have impaired psychomotor function (Table 2) for up to six days.74 Such effects are especially pronounced in older patients.75 The causative role of hypocapnia in postoperative cogni-tive dysfunction is underscored by the finding that exposure to an elevated partial pressure of arterial carbon dioxide during anesthesia appears to enhance postoperative neuropsychologic performance.75 Re-assuringly, according to studies of postoperative cog-nition in otherwise healthy patients, the adverse effects of hypocapnia, although often prolonged, appear to be reversible.
The exact role of hypocapnia in panic disorder is unclear. However, metabolic alkalosis induces panic in a substantial proportion of patients with panic disorder,76 and the central nervous system signs seen during panic attacks (e.g., dizziness, lightheadedness, confusion, and syncope) are consistent with the pres-ence of hypocapnia-induced cerebral hypoxia.77
Hypocapnia may be an important underlying mech-anistic link between panic disorder and other diseases. For example, patients with asthma and panic disorder
TABLE 2. ADVERSE NEUROLOGIC AND MYOCARDIAL
EFFECTS OF HYPOCAPNIA.
- Brain injury in neonates
- Multicystic encephalomalacia
- Cystic periventricular leukomalacia
- Pontosubicular necrosis
- Cerebral infarction
- Reactive hyperemia and hemorrhage Impairment of cerebral function in adults
- Increased time to regain consciousness, increased reaction times
- Poorer psychomotor performance, diminished higher intellectual functions
- Personality changes
- Myocardial effects
- Decreased myocardial oxygen supply
- Reduced coronary flow and collateral flow
- Increased coronary vascular resistance, increased
- risk of coronary-artery spasm Increased coronary microvascular leakage Increases in platelet count and aggregation
- Increased myocardial oxygen demand
- Increased oxygen extraction
- Increased (and later decreased) contractility
- Increased intracellular calcium concentration
- Increased systemic vascular resistance
- Myocardial ischemia
- Reperfusion injury
may be at increased risk for other illnesses.78 A majority of patients with recurrent chest pain but no angio-graphic evidence of coronary artery disease meet the diagnostic criteria for panic disorder. Because hypo-capnia is common in both of these groups, the possi-bility of underlying organic disease should always be considered in patients with hypocapnia.
Sudden exposure to very high altitude can result in long-term neurologic impairment. However, the central nervous system impairment seen in previous-ly healthy mountaineers after exposure to extremely high altitudes has been demonstrated to be most closely correlated with the degree of hypocapnia — not the level of hypoxia — attained.79 The cause of acute central nervous system symptoms at high alti-tudes appears to be alkalosis due to increased minute ventilation; such alkalosis can be prevented by pre-treatment with acetazolamide, which ameliorates the symptoms of high-altitude pulmonary edema.80
HYPOCAPNIA AND THE LUNG
Adverse pulmonary consequences of experimen-tally induced hypocapnia have been described in terms of effects on airways, alveolar-capillary perme-ability, lung compliance, and pulmonary vasculature, as well as the overall effect on lung injury.
Hypocapnia and the Tracheobronchial Tree
Airway hypocapnia increases airway resistance81 by inducing bronchospasm and increasing airway-micro-vasculature permeability (Fig. 4).82 Bronchoconstric-tion induced by hypocapnia may have adverse con-sequences.11,82 Although hypocapnia is a consistent feature of asthma, it is not clear whether it has a clin-ically important pathogenic role. More than 30 years ago, it was hypothesized that hypocapnia resulting from hyperventilation during an asthma attack may perpetuate the bronchospasm and culminate in a cy-cle of progressive hypocapnia and increasing bron-chospasm (Fig. 4).83 This theory is seldom discussed now, but considerable experimental evidence supports it.11,48,81,82 Furthermore, clinical data indicate that hypo-capnia may contribute to increased airway resistance in patients with asthma.11 In addition, alveolar hypocap-nia occurs during cardiopulmonary bypass, resulting in bronchoconstriction, increased airway resistance, and reduced lung compliance.84 These changes are reversed by the addition of inspired carbon dioxide.84
Acute Lung Injury
Aside from changes in airway resistance, hypocap-nia causes increased pulmonary-capillary permeabil-ity,53 parenchymal injury,85 and depletion of lamellar
Figure 4. Potential Role of Hypocapnia in Asthma.
Hypocapnia increases airway resistance by causing broncho-spasm and increased microvascular permeability. These ef-fects, in turn, increase the work of breathing and may potentiate the sensation of dyspnea, leading to further hyperventilation, progressive hypocapnia, and increasing bronchospasm, culmi-nating in a cycle of fatigue and respiratory failure.
bodies.86 These negative effects are all ameliorated by supplemental carbon dioxide.53,85,86 Hypocapnia decreases overall lung compliance in humans,87 per-haps because of effects on surfactant function. Finally, alveolar hypocapnia attenuates hypoxic pulmonary vasoconstriction, worsening intrapulmonary shunt and systemic oxygenation.49
Hyperventilation and hypocapnic alkalosis fre-quently coexist in patients with lung injury88; more-over, hyperventilation can cause acute lung injury. Al-though it is difficult to separate hyperventilation from hypocapnic alkalosis, the association of hyperventi-lation, hypocapnia, and worsened lung injury is in-
creasingly well documented.2,7,35 Such lung injury and related outcomes are conventionally considered to be due to excessive mechanical lung stretch. Thus, hypo-capnia is conventionally thought to play a passive — not a pathogenic — part in lung injury. However, the concept that hypocapnia might have a pathogen-ic role in the acute respiratory distress syndrome was first proposed in 1971 by Trimble and colleagues.41 They reported that, in a small study of patients with post-traumatic lung injury, hypocapnia was associat-ed with worsened pulmonary function that was re-versed by supplemental inspired carbon dioxide.41
Neonatal Lung Dysfunction
Both hyperventilation and hypocapnia have been identified as independent determinants of long-term pulmonary dysfunction in survivors of neonatal in-tensive care units.35 As noted above, hypocapnia is common in critically ill neonates and can potentiate many pathogenic lung processes; it is possible that hypocapnia may have a causative role in the develop-ment of bronchopulmonary dysplasia.35
HYPOCAPNIA AND THE
Cardiovascular effects of hypocapnic alkalosis in-clude alterations in myocardial oxygenation and car-diac rhythm (Table 2). In addition, hypocapnia may have a causal role in digital-artery spasm in periph-eral vascular disorders (e.g., Raynaud’s disease), pos-sibly, at least in part, because hypocapnic alkalosis causes or worsens vasoconstriction and enhances platelet aggregation.89
Acute hypocapnia decreases myocardial oxygen delivery while increasing oxygen demand (Table 2).90 Oxygen demand is increased through increases in my-ocardial contractility91 and systemic vascular resist-ance.92 In addition, hypocapnia may precipitate throm-bosis93 through increased platelet levels or platelet aggregation. These effects may contribute to the vari-ant angina that characteristically occurs with hyper-ventilation. Thus, hypocapnia may contribute to clin-ically relevant acute coronary syndromes.
Hypocapnia has been clearly linked to the devel-opment of arrhythmias, both in critically ill patients9 and in patients with panic disorder.77 Such effects may be secondary to ischemia, but specific direct myocar-dial effects may occur. Conversely, hypocapnic alkalosis may be therapeutically effective in arrhythmias in-duced by local anesthetics46 or tricyclic antidepres-sants94; in these cases, the alkalosis is the determinant of efficacy.
HYPOCAPNIA AND HEART–LUNG
Central sleep apnea results in hypoxemia, increased sympathetic nervous system activity, and daytime somnolence; when it occurs in patients with conges-tive heart failure, it increases the risk of death. An enhanced ventilatory response to carbon dioxide may contribute to the development of central sleep apnea in some patients with congestive heart failure,95 and hypocapnia triggers periodic respirations in these patients.96 One of the mechanisms by which applica-tion of noninvasive positive airway pressure reduces central sleep apnea is by increasing hemoglobin oxy-gen saturation and increasing the partial pressure of arterial carbon dioxide toward or above the apneic threshold.96 In fact, central sleep apnea is predicted by the presence of hypocapnia during waking hours.97 Thus, hypocapnia is a common finding in patients with sleep apnea and may be pathogenic.
HYPOCAPNIA AND HUMAN
In pregnant women, the partial pressure of arteri-al carbon dioxide is maintained at approximately 10
mm Hg lower than in nonpregnant women (Table 1). This physiologic state is associated with lowered serum bicarbonate-ion concentrations, which revert to normal values shortly after delivery. However, fur-ther lowering of the partial pressure of arterial carbon dioxide — even for a short duration, such as during anesthesia for cesarean section — may have serious adverse affects on the fetus (such as decreased fetal oxygen tension, increased base deficit, lower Apgar scores, and delayed onset of rhythmic neonatal breath-ing).98 These effects may be prevented by adminis-tering inspired carbon dioxide.98 Alkalosis associated with hypocapnia decreases placental perfusion, re-duces umbilical-vein oxygen tension,99 and causes reflex spasm of the umbilical vein.45 Because carbon dioxide increases fetal respiration, which may cause increased stretch and distention of the lung,100 fetal hypocapnia may impair pulmonary maturation.
Hypocapnia is neither a benign clinical entity nor an epiphenomenon. On the contrary, increasing ev-idence suggests that hypocapnia appears to induce substantial adverse physiological and medical effects. Thus, the decision to institute hypocapnia for thera-peutic purposes should be undertaken only after care-ful consideration of the risks and benefits and should in general be limited to emergency management of life-threatening increases in intracranial pressure or pulmonary-vascular resistance. The risk of accidental hypocapnia should be recognized and measures tak-
en to prevent it. Prophylactic induction of hypocap-nia currently has no clinical role.
Supported by the Canadian Institutes of Health Research. Dr. Kavanagh is the recipient of a Premier’s Research Excellence award from the Ontario Ministry of Energy, Science, and Technology, and Dr. Laffey is the recipient of a Training Fellowship from the Health Research Board, Ireland.