Effect of hypobaria on ventilatory and CO2 responses to short-term hypoxic exposure in cats

The effect of hypobaria on the ventilatory response to short-term hypoxia was studied by comparing the respiratory mechanical and inspired CO₂ ventilatory responses to hypobaric hypoxia (438 mmHg) with normobaric hypoxia (11.8% ). Fifteen spontaneously breathing, anesthetized cats were divided into three groups of five: time control, normobaric hypoxia and hypobaric hypoxia. Measurements of ventilation, gas exchange, and responses to intermittent CO₂ rebreathing were collected over a 4 h period. Pa_O2 fell to 44.5±2.7 mmHg, Pa_CO2 fell to 24.8±0.9, and pH rose to 7.49±0.01 in both hypoxic groups. Tidal volume did not change with respect to time or condition, but frequency and ventilation were significantly increased in the hypobaric hypoxic group. The slope of the CO₂ response was unchanged over time or by condition. These results suggest that hypobaric hypoxia may alter the pattern of breathing responses to hypoxia but not the CO₂-response. If metabolic rate remained constant, these results could be explained by a difference in dead space between hypoxic conditions.     

Lack of positive interaction between CO2 and hypoxic stimulation for P(CO2)-VAS response slope in humans

To compare the effect of hypoxia on ventilatory responses and respiratory sensation to carbon dioxide, 29 young adults were examined using a modified Read's rebreathing method with four experimental conditions. We used varying gas mixtures and kept PET(O2) constant at >300, 100, 80 and 60 mmHg for each four rebreathing tests. Respiratory sensation was measured by visual analog scale (VAS). The slope of the CO2-ventilation response curve increased significantly with hypoxia, confirming a positive ventilatory interaction between hypoxia and hypercapnia. However, the slope of the CO2-VAS response curve remained unchanged. The V(E)-VAS relation slope tended to become depressed with advancing hypoxia, i.e. the magnitude of VAS elicited by a given ventilation decreased with hypoxia, signifying that dyspneic sensation was effectively mitigated during hypoxic hyperventilation. We suggest that this relief of dyspneic sensation might be due to the inhibitory respiratory effect from augmented pulmonary stretch receptor (PSR) activity.     

Alveolar hypoventilation syndrome. Studies of ventilatory control in patients selected for diaphragm pacing

The syndrome of nocturnal hypoventilation and/or apnea without major intrinsic lung disease is of greater surgical interest since the introduction of therapeutic diaphragm pacing. Of our 29 patients successfully managed by diaphragm pacing, we describe 17 in whom detailed studies of the ventilatory responses to hypoxia are available.The diagnosis was established by the following criteria: (1) an increase in arterial carbon dioxide tension (PaCO₂) (> 45 mm Hg) and a depression in the arterial oxygen tension (PaO₂) (< 75 mm Hg) at least during sleep; (2) respiratory arrests and/or apneic episodes during sleep; (3) near normal tests of ventilatory capacity; and (4) diminished ventilatory response to carbon dioxide breathing.New information on the ventilatory responses to hypoxia was obtained by a nonsteady state closed-circuit rebreathing method. The following were compared: (1) hypoxia with carbon dioxide variably absorbed so as to maintain base line PaCO₂ (“normo”-capnic hypoxia); (2) 50 per cent oxygen with carbon dioxide accumulation (hypercapnia hyperoxia); and (3) carbon dioxide accumulation in room air (hypercapnic hypoxia). The data indicate (1) 15 patients showed no measurable ventilatory response to “normo”-capnic hypoxia and two patients showed blunted responses; (2) seven patients showed no ventilatory responses to carbon dioxide whereas 10 patients showed suboptimal increases; (3) combined hypoxia and hypercapnia caused a suboptimal ventilatory response in four patients. These same four patients showed blunted as opposed to absent ventilatory responses to carbon dioxide. (4) Five patients showed no ventilatory responses to all combinations.These 17 patients together with three others not so extensively studied have been managed by nocturnal diaphragm pacing for periods of four months to nearly eight years.We conclude (1) ablation or severe blunting of the ventilatory responses to hypoxia is a frequent and previously underestimated component of this disorder; and (2) diaphragm pacing is an appropriate form of therapy.     

Acute respiratory failure and obesity with normal ventilatory response to carbon dioxide and absent hypoxic ventilatory drive

We measured hypoxic and hypercapnic ventilatory drive in a 64 year old woman with acute respiratory failure, congestive heart failure and obesity when she was in remission. She had a ventilatory response to carbon dioxide (CO₂) comparable to that in six obese women without hypoventilation but no ventilatory response to hypoxia or to vital capacity breaths of 15 per cent CO₂ in N₂- Following weight loss, her ventilatory response to CO₂ increased but hypoxic ventilatory drive remained absent. These findings indicate that attenuation of hypoxic ventilatory drive caused by loss of peripheral chemoreceptor function can be a predisposing factor in the development of acute respiratory failure associated with obesity.     

Normal chemoreceptor function in obesity before and after heal bypass surgery to force weight reduction

Ventilatory responses to progressive isocapnic hypoxia and rebreathing of carbon dioxide in oxygen were determined in four obese women before and approximately 1 year after ileal bypass surgery to force weight reduction. None of the patients was hypoventilating and all had normal pulmonary function tests. The ventilatory responses to hypoxia were normal before surgery and were not effected by weight reduction. The ventilatory responses to hypercapnia did not change in slope but a shift of the carbon dioxide response line toward a lower arterial carbon dioxide tension occurred in two subjects after weight reduction. We conclude that obesity per se does not necessarily cause loss of hypoxic ventilatory drive.     

Ventilatory response to CO2 and hypoxia during cold exposure in awake rats

Recently, we have described the effects of hypoxia and of hypercapnia on the metabolic (V̇_O2) and ventilatory responses to cold in unanesthetized intact and carotid body-denervated (CBD) rats (Gautier et al., J. Appl. Physiol. 73: 847–854, 1992 and 75: 2570–2579, 1993). In the present paper, we have reanalyzed the above results for a more detailed study of the interactions of hypoxia (Fi_O2 = 0.12), hypercapnia (Fi_CO2 = 0.04) and changes in V̇_O2 with the ventilatory control. The results show that: (1) Compared to normoxia, in hypoxia increments in V̇ and Vt are proportional to V̇_O2 whereas in hypercapnia increments in ventilation (V̇) and tidal volume (Vt) are independent of V̇_O2. In both hypoxia and hypercapnia, increases in respiratory frequency (fr) are independent of V̇_O2; and (2) Interactions of hypoxia, hypercapnia and V̇_O2 with control of V̇ persist in CBD rats but, for a given V̇_O2, V̇, Vt and fr are lower than in intact rats. These interactions are essentially similar to those observed during muscular exercise performed in normoxia, hypoxia or hypercapnia. It is suggested that during cold exposure or muscular exercise, resulting both in increased V̇_O2, there are common integrative structures probably located in the hypothalamus which are involved in the control of breathing.     

Effect of Carbon Dioxide on Pulmonary Vascular Tone at Various Pulmonary Arterial Pressure Levels Induced by Endothelin-1

There have been contradictory reports suggesting that CO₂ may constrict, dilate, or have no effect on pulmonary vessels. Permissive hypercapnia has become a widely adopted ventilatory technique used to avoid ventilator-induced lung injury, particularly in patients with acute respiratory distress syndrome (ARDS). On the other hand, respiratory alkalosis produced by mechanically induced hyperventilation is the mainstay of treatment for newborn infants with persistent pulmonary hypertension. It is important to clarify the vasomotor effect of CO₂ on pulmonary circulation in order to better evaluate the strategies of mechanical ventilation in intensive care. In the present study, pulmonary vascular responses to CO₂ were observed in isolated rat lungs (n = 32) under different levels of pulmonary arterial pressure (PAP) induced by various doses of endothelin-1 (ET-1). The purposes of this study were to investigate (1) the vasodilatory effect of 5% CO₂ in either N₂ (hypoxic-hypercapnia) or air (normoxic-hypercapnia) at different PAP levels induced by various doses of endothelin-1, and (2) the role of nitric oxide (NO) in mediating the pulmonary vascular response to hypercapnia, hypoxia, and ET-1. The results indicated that (1) CO₂ produces pulmonary vasodilatation at high PAP under ET-1 and hypoxic vasoconstriction; (2) the vasodilatory effect of CO₂ at different pressure levels varies in accordance with the levels of PAP, the dilatory effect tends to be more evident at higher PAP; and (3) endogenous NO attenuates ET-1 and hypoxic pulmonary vasoconstriction but does not augment the CO₂-induced vasodilatation.     

Possible role of dopamine in ventilatory acclimatization to high altitude

Ventilatory acclimatization to high altitude is accompanied by increased hypoxic (HVR) and hypercapnic (HCVR) ventilatory responses which may reflect increased carotid body chemosensitivity. Dopamine is an inhibitory neuromodulator of the carotid body and its activity may be reduced by hypoxic exposure. To determine whether decreased dopaminergic activity could account for the increased chemosensitivity of acclimatization, we examined the response to peripheral dopamine receptor (D₂) blockade with domperidone on HVR and HCVR in awake cats before and after exposure to simulated altitude of 14000 ft for 2 days. During anesthesia, we also examined the effects of domperidone on carotid body responses to hypoxia and hypercapnia in acclimatized and low altitude cats. Two days' exposure to hypobaric hypoxia produced an increase in HVR and HCVR. Before acclimatization, domperidone augmented HVR and HCVR, but there was no effect after acclimatization. In anesthetized low altitude cats, domperidone increased carotid body responses to hypoxia and hypercapnia, but had no effect in acclimatized cats. These results indicate that decreased endogenous dopaminergic activity may contribute to increased ventilatory and chemoreceptor responsiveness to hypoxia and hypercapnia during hypoxic ventilatory acclimatization.     

Hypercapnic Ventilatory Control in Patients with Chronic Airflow Obstruction: A Follow-up Study*

Abstract: Hypercapnic ventilatory control in patients with chronic airflow obstruction; a follow-up study. P. V. Zimmerman, N. Maranetra and M. C. F. Pain. Aust. N.Z. J. Med., 1982, 12 , pp. 504–510. A four year follow-up study of ventilatory control has been performed in 50 patients with chronic airflow obstruction. Hypercapnia was induced by rebreathing, and both the ventilation and the mouth pressures produced during transient airway conclusion during inspiration were obtained. Both the chemoreceptor sensitivity to hypercapnia (AP/APCO₂) and the ventilatory response (AV/APCO?) measured during the initial study were shown to be unrelated to prognosis when the outcome of these patients was determined four years later. A relationship between initial hypercapnia and survival was found. Twenty of the patients were restudied. No significant change occurred in the chemosensitivity to hypercapnia over the four year period. In the initial study and the follow-up study hypercapnia was also found to be associated with a reduced chemoreceptor sensitivity. Although both chemosensitivity to hypercapnia and survival were found to be separately related to arterial CO₂ tension, there was no direct relationship between them. These results suggest that any possible influence of chemosensitivity on survival is likely to be overshadowed by other factors.     

The ventilatory and metabolic response to hypercapnia in newborn mammalian species

Conscious newborns of 12 species from 4 mammalian orders, ranging in body mass (M) from 1 g (mouse) to 5 kg (deer), were studied during air and during 5% CO₂ breathing. The interspecies relationship between oxygen consumption (V̇_O2) and M was the same in air and hypercapnia, in both cases V̇_O2 ∝ M^0.90; on average, hypercapnic V̇_O2 was 101% of the air value. In 5% CO₂, ventilation (V̇e) increased in all newborns, mostly because of the increase in tidal volume (178%), whereas breathing rates averaged 98% of the air values. The hyperpnea during CO₂ was slightly greater in the larger newborns. Body temperature was not altered by CO₂ breathing. We conclude that the average respiratory response of the newborn to moderate hypercapnia is a hyperventilation different from that of the neonatal mammal in acute hypoxia (Mortola et al., Respir. Physiol. 78: 31–43, 1989). In fact, hypercapnic hyperventilation resulted only from the hyperpnea, with no hypometabolic contribution, and the hyperpnea reflected the increase in tidal volume, with no change in rate. It is also concluded that the neonatal hypometabolic response is specific to hypoxia, and not an undifferentiated response to chemoreceptors stimulation.     

The biphasic ventilatory response to hypoxia in preterm infants is not due to a decrease in metabolism

The mechanism underlying the biphasic ventilatory response to hypoxia in neonates is poorly understood. Because alveolar P_CO2 (P_ACO2) decreases and remains low during hypoxia, it has been argued that a decrease in metabolism may occur. We hypothesized that if the late decrease in ventilation during hypoxia is due to a decrease in CO₂ production, an increase in P_ACO2 should abolish it. We studied 27 preterm infants [birth weight, 1,700 ± 41 g (mean ± SEM); study weight, 1,760 ± 36 g; gestational age 32 ± 0.2 weeks; postnatal age, 17 ± 1 days]. A flow-through system and Beckman analyzers were used to measure ventilation and alveolar gases. Metabolism was expressed as changes in oxygen consumption. Infants were studied randomly during hypoxia alone (15% O₂ + N₂, n = 55) and during hypoxia plus CO₂, (0.5% CO₂, n = 30; 2% CO₂, n = 10). Each experiment consisted of 2 minutes of control measurements (21% O₂), 5 minutes of measurements during hypoxia alone or hypoxia plus CO₂, followed by 2 minutes of recovery (21% O₂). We found a biphasic response to hypoxia with or without CO₂ supplementation, the percent change in ventilation from initial peak hyperventilation to late hypoventilation at 5 minutes being -16 ± 2 on 15% O₂; -9 ± 3 on 15% O₂; + 0.5% CO₂ and -15 ± 9 on 15% O₂; + 2% CO₂; (P < 0.05).The decrease in ventilation was primarily due to a significant decrease in frequency; tidal volume increased. Oxygen consumption decreased similarly with the various inspired gas mixtures during hypoxia. These findings indicate that the decrease in ventilation during hypoxia is unlikely to be solely due to a decrease in metabolism since the late decrease in ventilation following initial hyperventilation still occurred despite the elimination of a fall in P_ACO2. We speculate that the mechanism underlying the late decrease in ventilation is likely of central origin, probably mediated through the release of inhibitory neurotransmitters. Pediatr Pulmonol. 1996; 22:287–294. © 1996 Wiley-Liss, Inc.     

Circulating endogenous opioids and ventilatory response to CO2 and hypoxia

The role of endogenous opioids in the control of breathing is not yet well defined. Radioimmunoassays that measure beta-endorphin (BE) and met-enkephalin (MET) having recently become available, we decided to evaluate the possible relation between the blood levels of these two opioids and different hypercapnic and hypoxic ventilatory responses observed in a group of normal subjects. Ventilatory response to hypercapnia (n = 9) and to hypoxia (n = 7) were determined by classical rebreathing methods. A voluntary isocapnic normoxic hyperventilation test was used as a control. Basal levels of BE and MET did not correlate with the magnitude of the ventilatory response to either hypercapnia or hypoxia. Moreover, BE and MET levels measured repeatedly up to 30 min after each test did not change significantly. We conclude that circulating endogenous opioids do not play a role in the control of breathing in normal humans. These results do not rule out a possible role for these substances as locally released mediators.     

Aging effects on the interaction of hypercapnia and hypoxia as ventilatory stimuli

We measured ventilatory responses to progressive hypercapnia at two steady-state levels of oxygenation and to progressive hypoxia at two steady-state levels of CO2 in 10 elderly and 10 young individuals. Under hyperoxic conditions, the ventilatory response to progressive hypercapnia was not significantly different between age groups but, under hypoxic conditions, the response to hypercapnia was lower in the elderly group. The interaction of hypercapnic and hypoxic stimuli was greater among young persons as indicated by a higher ratio of the hypercapnic response slopes (hypoxic/hyperoxic); 1.48 +/- 0.19 versus 0.98 +/- 0.11, p less than .05. The ventilatory response to hypoxia at the lower CO2 level was significantly greater among elderly than among young adults but not significantly different between age groups at the higher CO2 level. The ratio of hypoxic response slopes (high PCO2/lower PCO2) was 1.56 +/- 0.17 among elderly participants and 3.14 +/- 0.63 among young participants (p less than .05). These results suggest that aging diminishes the multiplicative effect of hypercapnia and hypoxia as ventilatory stimuli.     

Endogenous opiates and ventilatory acclimatization to chronic hypoxia in the cat

The effects of the opiate antagonist naloxone (0.4 mg.kg-1, i.v.) on carotid chemoreceptor and ventilatory responses to graded steady-state levels of hypoxia and hypercapnia were investigated in two groups of cats: chronically normoxic and chronically hypoxic. The cats of the latter group were exposed to PIO2 of about 70 mm Hg at sea level for 3-4 weeks and showed an attenuated response to hypoxia. All cats were tested under alpha-chloralose anesthesia. Naloxone treatment did not increase appreciably carotid chemoreceptor activity or its responses to hypoxia and hypercapnia in either cat group. Naloxone caused a small ventilatory stimulation in the chronically hypoxic cats, so that the attenuated response to hypoxia was not relieved. By contrast, the chemoreflex ventilatory response to hypoxia was stimulated by naloxone in the chronically normoxic cats. The findings that the depressed ventilatory chemoreflexes in the chronically hypoxic cat were not ameliorated by the opiate antagonist indicate that an increased elaboration of endogenous opiates does not underlie ventilatory adaptation to chronic hypoxia.     

Hypoxic and hypercapnic ventilatory responses in awake children with congenital central hypoventilation syndrome

Congenital central hypoventilation syndrome (CCHS) has been thought to be a disorder of central chemoreceptor responsiveness. Previous studies in CCHS have shown decreased or absent ventilatory responsiveness to both hypercarbia and hypoxia. However, hypoxic responsiveness during wakefulness has not been systematically studied. We studied hypoxic and hypercapnic ventilatory responses during wakefulness in five children with CCHS (6 to 11 yr of age). To measure the hypercapnic response, the children rebreathed a hyperoxic hypercapnic mixture until PaCO2 reached 56 to 69 mm Hg. For the hypoxic response, the children rebreathed a hypoxic gas mixture, at mixed venous PCO2, until SaO2 had fallen to less than 78%. We found that the ventilatory responses to hypercapnia and hypoxia were very variable (linear correlation coefficients ranging from -0.44 to +0.63 for hypercapnic responses and from -0.15 to +0.77 for hypoxic responses), with no significant change from baseline in response to either stimulus. There was no evidence of progressive ventilatory stimulation despite increasing stimulus. Additionally, these children had no subjective sensation of dyspnea or discomfort. This establishes that hypoxic and hypercapnic ventilatory control is absent during wakefulness. Chemoreceptor control (peripheral and central) is, therefore, defective in all states in children with CCHS. We speculate that the defect in CCHS lies in central integration of the central and peripheral chemoreceptor signals.     

Plasma Adenosine during Investigation of Hypoxic Ventilatory Response

Adenosine, an endogenous nucleoside, is released by hypoxic tissue, causes vasodilation, and influences ventilation. Its effects are mediated by P1-purinoceptors. We examined to what extent the plasma adenosine concentration in the peripheral venous blood correlates with hypoxic ventilatory response (HVR) and ventilatory drive P0.1 to find out whether endogenously formed adenosine has an influence on the individual ventilatory drive under hypoxic conditions. While investigating the HVR of 14 healthy subjects, the ventilatory drive P0.1 was measured with the shutter of a spirometer. Determination of the ventilatory drive P0.1(RA) started under room air conditions (21% O₂) and then inspiratory gas was changed to a hypoxic mixture of 10% O₂in N₂to determine P0.1(Hyp). At the time of the P0.1 measurements, two blood samples were taken to determine the adenosine concentrations. After removal of cellular components and proteins, samples were analyzed by high-pressure liquid chromatography (HPLC). Both adenosine concentrations in plasma under room air (r = 0.59, p< 0.05) and adenosine concentrations under hypoxia (r = 0.75, p< 0.01) correlated significantly with the ventilatory drive P0.1. In addition, plasma adenosine concentrations during hypoxic conditions showed a significant correlation with HVR on the 0.01 level (r = 0.71, p< 0.01). The results indicate a possible role of endogenous adenosine in the regulation of breathing in humans. We assume that endogenous adenosine influences the HVR and the ventilatory drive, probably by modulating the carotid body chemoreceptor response to hypoxia.     

Respiratory chemosensitivity in smokers. Studies on monozygotic twins

Whether smoking habits influence ventilatory response to hypoxia and hypercapnia was examined in 23 pairs (8 concordant smoker pairs, 8 concordant nonsmoker pairs, and 7 discordant pairs) of monozygotic twins. Pulmonary function tests were also performed. Smokers (7 subjects) were significantly taller by 2 cm (mean value) than nonsmoker partners (7 subjects). Arterial pH of smokers was significantly lower, although it stayed within normal range. Closing volume to vital capacity ratio tended to be larger in smokers. Ventilatory response to hypoxia (slope factors for ventilation-alveolar Po2 curve and ventilation-arterial O2 saturation line) was significantly higher in smoking than in nonsmoking partners, whereas ventilatory response to hypercapnia (slope factor for ventilation-alveolar Pco2 line) was not different. Intrapair variances for height and ventilatory response to hypoxia were larger in discordant pairs with respect to smoking than in concordant pairs. These results indicate that smoking habits influence the slope, without the parallel shift, of the hypoxic ventilatory response curve.     

Sleep deprivation and the control of ventilation

Sleep deprivation is common in acutely ill patients because of their underlying disease and can be compounded by aggressive medical care. While sleep deprivation has been shown to produce a number of psychological and physiologic events, the effects on respiration have been minimally evaluated. We therefore studied resting ventilation and ventilatory responses to hypoxia and hypercapnia before and after 24 h of sleeplessness in 13 healthy men. Hypoxic ventilatory responses (HVR) were measured during progressive isocapnic hypoxia, and hypercapnic ventilatory responses (HCVR) were measured using a rebreathing technique. Measures of resting ventilation, i.e., minute ventilation, tidal volume, arterial oxygen saturation, and end-tidal gas concentrations, did not change with short-term sleep deprivation. Both HVR and HCVR, however, decreased significantly after a single night without sleep. The mean hypoxic response decreased 29% from a slope of 1.20 +/- 0.22 (SEM) to 0.85 +/- 0.15 L/min/% saturation (p less than 0.02), and the slope of the HCVR decreased 24% from 2.07 +/- 0.17 to 1.57 +/- 0.15 L/min/mmHg PCO2 (p less than 0.01). These data indicate that ventilatory chemosensitivity may be substantially attenuated by even short-term sleep deprivation. This absence of sleep could therefore contribute to hypoventilation in acutely ill patients.     

Are bedding and rebreathing suffocation a cause of SIDS?

Suffocation by bedclothes became a popular diagnosis in the 1940s but gradually became replaced with the diagnostic label of Sudden Infant Death Syndrome (SIDS). In 1991 a paper purported that, instead of SIDS, pillows filled with polystyrene beads had caused death by rebreathing suffocation; this conclusion was reached on the basis of experiments with anesthetized rabbits breathing through a doll's head that was placed face down on the pillow. Because of the anesthesia, rabbits could not change their face down position. The doll's nares could not collapse, which would have resulted in rapid death due to conventional suffocation. The rabbits required up to 3 hours or more to die of hypercarbia and hypoxia. Studies in normal infants revealed that they turned from the face-down position after only 2 minutes. (The only infant who retained CO₂ soon died of a fatal neurologic disorder, with central hypoventilation). Using the rabbit/doll's head and mechanical models, a wide range of bedding was indicted, including cushions, sheepskins, pillows, comforters, foam mattresses, and even simple blankets and sheets as potentially causing fatal rebreathing. Except for the use of pillows in general, as well as mattresses filled with kapok and bark, there has been no epidemiologic support for these indictments. Although normal infants are unlikely to succumb to rebreathing suffocation, infants with blunted ventilatory responsiveness and delayed arousal duo to prior hypoxia were hypothesized to be at increased risk. Support for this concept was found in the pathology of the brain stem in victims of SIDS that was attributed to prior hypoxic injury. In infants who survived prolonged apnea, less than 20% have demonstrated a diminished ventilatory responsiveness to hypercarbia, but, more significantly, none had an absent response. Arousal to hypercarbia, an abnormality which is crucial to the hypothesis of rebreathing suffocation, is regularly present in normal subjects, but the threshold is higher in near-SIDS infants; however, no instances of failure to arouse have been reported in near-SIDS. If the infant is placed on his or her back or side, the issue of bedding could become moot; unfortunately, a sizable percentage of infants are still being placed prone for sleep. Instead of confusing parents with an ever-expanding list of “dangerous bedding,” the message “Back to Sleep” should be emphasized. Pediatr Pulmonol, 1996; 22:335–341. © 1996 Wiley-Liss, Inc.     

Depression of peripheral chemosensitivity by a dopaminergic mechanism in patients with obstructive sleep apnoea syndrome.

In the present study, respiratory drives to chemical stimuli and peripheral chemosensitivity were evaluated in patients with obstructive sleep apnoea (OSAS). The effects of oral administration of domperidone, a selective dopamine D2-receptor antagonist, were also examined, to study the respiratory effects of endogenous dopamine on peripheral chemoreceptors. Sixteen patients with OSAS and nine normal control subjects were studied. Respiratory responses to hypercapnia and hypoxia were measured using the rebreathing method and isocapnic progressive hypoxia method, respectively. The hypoxic withdrawal test, which measures the decrease in ventilation caused by two breaths of 100% O2 under mild hypercapnic hypoxic conditions (end-tidal oxygen and carbon dioxide tensions approximately 8.0 kPa and 5.3-6.7 kPa, respectively), was used to evaluate peripheral chemosensitivity. In the patients with OSAS, ventilatory responses to hypercapnia and hypoxia were significantly decreased compared with those of control subjects. Hypoxic withdrawal tests showed that peripheral chemosensitivity was significantly lower in patients with OSAS than in normal subjects. Hypercapnic ventilatory response and peripheral chemosensitivity were enhanced by administration of domperidone in the patients with OSAS, although no changes in either of these were observed in the control subjects. The hypoxic ventilatory response and peripheral chemosensitivity in the patients with OSAS were each significantly correlated with severity of hypoxia during sleep. These findings suggest that peripheral chemosensitivity in patients with obstructive sleep apnoea syndrome may be decreased as a result of abnormality in dopaminergic mechanisms and that the reduced chemosensitivity observed in patients with obstructive sleep apnoea syndrome may affect the severity of hypoxia during sleep.