Retinal Arteriolar and Middle Cerebral Artery Responses to Combined Hypercarbic/Hyperoxic Stimuli

Abstract

purpose. The relative effect of simultaneously administered oxygen and carbon dioxide on the retinal and cerebral vessels is still controversial. The purpose of this study was to quantify and compare the superior-temporal retinal arteriole (RA) and middle cerebral artery (MCA) responses to hypercarbic and combined hypercarbic/hyperoxic stimuli.

methods. Twelve young, healthy volunteers participated in the study. End-tidal pressure of carbon dioxide was raised and maintained at 22% from baseline (hypercarbia), while end-tidal pressures of oxygen (P_ETO₂) of 100 (normoxia), 500, and 300 mm Hg (hyperoxia) were instituted. RA diameter and blood velocity were measured with laser Doppler velocimetry and simultaneous vessel densitometry; MCA blood velocity was measured with transcranial Doppler ultrasound.

results. Normoxic hypercarbia increased RA blood velocity by +17% and calculated flow by +21%. Hypercarbia/hyperoxia-500 mm Hg decreased RA diameter by −8%, velocity by −16% and calculated flow by −29%. MCA blood velocity increased by +45% in response to normoxic hypercarbia, significantly greater than RA blood velocity (P < 0.001). Increase in P_ETO₂ did not affect the hypercarbia-induced increase in MCA blood velocity.

conclusions. Hyperoxia reversed hypercarbia-induced vasodilation in RA in a concentration-dependent manner. Hypercarbia induced greater vasodilation in the MCA than in the RA but MCA blood velocity was unaffected by increases in P_ETO₂.

Dysregulation of ocular blood flow plays a prominent role in the pathogenesis of many ocular diseases. ^{1 2} Alterations of blood flow in intraocular, retrobulbar, and cerebral vessels have been reported in patients with diabetes, ^{3 4 5 6} glaucoma, ⁷ and Behçet’s disease with ocular involvement. ⁸ Moreover, it is generally accepted that the status of the retinal vessels not only serves as a predictor of retinal disease development, but also is an indicator of cerebrovascular health. Abnormal retinal vessel calibers are associated with cerebrovascular disease, an increased risk of stroke, and lower brain oxygenation. ^{9 10 11 12} Correlation between decreased cerebral and retinal vascular reactivity were found in patients with cerebral small vessel disease. ¹³ Innovative imaging techniques to assess ocular hemodynamics have contributed to the understanding of the role of the vasculature in the pathophysiology of ocular vascular diseases (see Ref. ¹⁴ for review). Homeostatic ocular blood flow measurements exhibit large intersubject variability. Consequently, provocative stimuli (vasoconstrictor or vasodilator) have been used to quantify vascular reactivity, which shows greater consistency. ¹⁵ In the eye, oxygen (O₂) is a potent vasoconstrictor, ¹⁶ whereas carbon dioxide (CO₂) causes vasodilation. ^{17 18} Retinal vessels show greater response to O₂, whereas choroidal and cerebral vascular beds respond more to CO₂. ^{17 19} Controversy still exists, however, in regard to the combined effect of O₂ and CO₂ on these vascular beds (see Ref. ¹⁶ for review). The conflicting conclusions can be partly explained by the interpretation of data from different techniques used to assess blood flow as well as differing methodologies used to provoke vascular reactivity. We have shown that concentration-dependent hyperoxia-induced vasoconstriction predominates over hypercarbia-induced vasodilation in the retinal arterioles. ²⁰ The purpose of this study was to quantify and compare the vascular responses of retinal arterioles and the middle cerebral artery to a series of standardized vasoactive stimuli. 

Methods

The study was approved by the Research Ethics Board of the University Heath Network, University of Toronto and adhered to the guidelines of the Declaration of Helsinki. Informed consent was obtained from each subject. Twelve healthy nonsmokers (1 woman) of mean age 25 years (SD ±5) were recruited into the study. All subjects had a refractive error less than ±6.00 D sphere and/or ±2.50 D cylinder, no ocular or systemic diseases, and no medications. 

The assessment of retinal and cerebral vascular reactivity was performed with identical inhaled gas provocation protocols on two separate days. Subjects sat undisturbed for 20 minutes before beginning the protocol. All subjects breathed via a commercial sequential gas delivery breathing circuit (HiOx-80; VIASYS Healthcare Inc., Dublin, OH) modified by adding a rebreathing bag to the expiratory port. ^{18 21} Adhesive tape (Tegaderm; 3M Health Care, St. Paul, MN) was used to ensure an airtight seal of the mask to the face. Gas was sampled continuously from inside the mask and analyzed for end-tidal partial pressures of CO₂ (P_ETCO₂) and O₂ (P_ETO₂). Noninvasive systolic and diastolic arterial blood pressures (at 1 minute intervals), and blood oxygen saturation and heart rate were monitored continuously by pulse oximetry (Cardiocap/5; Datex-Ohmeda, Tewksbury, MA). The method of gas administration and its advantages have been published. ^{22 23} Briefly, targeted end-tidal gas concentrations (partial pressures), as described in the protocol were preprogrammed and implemented with a custom-built computer-controlled automated gas blender and sequencer (RespirAct; Thornhill Research Inc., Toronto, ON, Canada). 

Procedures

The breathing protocol is illustrated in Figure 1 . Each stage of the experiment was performed when P_ETCO₂ and P_ETO₂ stabilized (i.e., <2 mm Hg change over a 2-minute period). After stabilization at baseline values of P_ETCO₂ and P_ETO₂ (normocarbia/normoxia I), P_ETCO₂ was then targeted to achieve a 20% increase from the baseline at a P_ETO₂ of 100 mm Hg (hypercarbia/normoxia I). P_ETO₂ was then increased to target 500 mm Hg (hypercarbia/hyperoxia-500) and subsequently decreased to normoxia, while maintaining P_ETCO₂ constant at 20% above baseline (hypercarbia/normoxia II). Next, P_ETCO₂ was decreased to baseline (normocarbia/normoxia II). Finally, both P_ETCO₂ and P_ETO₂ were increased simultaneously to achieve 20% increase in P_ETCO₂ and 300 mm Hg P_ETO₂ (hypercarbia/hyperoxia-300). Each condition was maintained for 5 minutes after targeted levels of P_ETCO₂ and P_ETO₂ were achieved and had stabilized. Our approach differed cardinally from previously published studies in that, by implementing continuous end-tidal gas monitoring, we ensured that all blood flow measurements were taken when similar respiratory parameters were achieved and confirmed across subjects. The overall time of each experimental condition for every subject varied from 8 to 15 minutes, depending on the length of transition phase when no blood flow measurements were taken. 

Quantitative Assessment of Retinal Blood Flow.

The pupil of the study eye was dilated with 1 drop of tropicamide 1%. Retinal blood flow was assessed with laser Doppler velocimetry and simultaneous vessel densitometry (Laser Blood Flowmeter, CLBF, model 100; Canon, Tokyo, Japan) in the superior-temporal arteriole, approximately 1 disc diameter from the optic nerve head, in a straight vessel segment distant from bifurcations. The instrument and measurement site selection details have been described previously. ²⁰ With this technique, retinal blood flow was calculated from measured diameter and velocity values. All measurements were made by a single experienced observer (MK). The subjects were masked to the breathing gas mixture composition. Retinal blood flow measurements were made for each condition only when end-tidal gas concentrations were stable. The quality of velocity waveforms was assessed against agreed-upon standards but the observer was masked to the quantitative results of the reading. Readings with loss of fixation and/or aberrant velocity waveforms were deleted at the time of measurement. The measurement of diameter was saved only if the coefficient of variation was less than 2%. At least 10 measurements were taken for each experimental condition. 

Quantitative Assessment of Cerebral Blood Flow.

Cerebral blood velocity was measured in the MCA with a 2-MHz pulsed Doppler ultrasound system (Multidop X4; DWL Elektronische System GmbH, Sipplingen, Germany). The MCA was identified by using an insonation pathway through the right or left temporal window just above the zygomatic arch. The Doppler signal was optimized by varying the depth (45–55 mm) and angle of insonation. ²⁴ When the optimal signal was obtained, the probe was fixed in place with a headband. Velocity (time averaged maximum or V _mean) was monitored continuously. 

Statistical Analysis

Statistical analysis was performed with commercial software (Statistica ver. 6; StatSoft, Tulsa, OK). All data are presented as the mean ± SD for each experimental condition. Repeated measures ANOVA was performed to test for any change in respiratory, systemic hemodynamic, retinal or cerebral hemodynamic parameters across experimental conditions and study visits. When appropriate, Tukey HSD post hoc tests were undertaken to determine the significance of any change relative to baseline. Tests of average correlations between arterial pressure parameters and retinal and cerebral hemodynamic parameters were used to assess any influence of systemic blood pressure on retinal and cerebral blood flows. 

Results

Twelve subjects completed retinal hemodynamic assessment, and their results were included in the analysis of retinal vascular reactivity. Ten subjects completed cerebral hemodynamic assessment, whereas two subjects did not have a temporal bone window that provided a good-quality Doppler signal. The data of these two subjects were excluded from the comparison of retinal and cerebral vascular reactivity. 

Respiratory Parameters

A stable increase in P_ETCO₂ was achieved throughout the experiment: +8.9 mm Hg (SD ±1) during assessment of retinal vascular reactivity and +9.8 mm Hg (SD ±1) during assessment of cerebral vascular reactivity. Attained P_ETCO₂ and P_ETO₂ values for each condition did not differ between the two study days (Table 1) . 

Systemic Hemodynamic Parameters

Systemic hemodynamic changes are shown in Table 2 . The differences in systolic (SP), diastolic (DP), and mean arterial blood pressure (MAP) between the first baseline and hyperoxia-500 were small and reached statistical significance only for SP (P < 0.001) and DP (P < 0.05). The differences in SP, DP, and MAP between the first baseline and hyperoxia-300 were larger than those between the first baseline and hyperoxia-500 and in the former reached statistical significance for SP, DP, and MAP (P < 0.05). This effect was consistent across the two visits. Systolic blood pressure was slightly higher on the second study day (P < 0.05). Mean arterial pressure was not different on the two study days. 

Superior-Temporal Retinal Arteriolar Vascular Reactivity

RA diameter did not change in response to hypercarbia/normoxia I (115 ± 14 μm vs. 113 ± 13 μm, P = 0.99). RA diameter decreased to 105 ± 14 μm (P < 0.001) in response to hypercarbia/hyperoxia-500. RA diameter returned to baseline during hypercarbia/normoxia II and during normocarbia/normoxia II (112 ± 12 and 112 ±13 μm, respectively) and then decreased to 107 ± 12 μm (P < 0.05) during hypercarbia/hyperoxia-300 (Fig. 2) . 

RA blood velocity increased in response to hypercarbia/normoxia I and decreased during hypercarbia/hyperoxia-500 (from baseline 30 ± 6 to 35 ± 6 mm/s, P < 0.01, and 25 ± 7 mm/s, P < 0.05, respectively). Velocity returned to baseline levels during hypercarbia/normoxia II, normocarbia/normoxia II, and hypercarbia/hyperoxia-300 (Fig. 3) . 

Calculated RA blood flow increased in response to hypercarbia/normoxia I and then decreased during hypercarbia/hyperoxia-500 (from baseline 9.2 ± 2.6 μL/min to 11.1 ± 3.5 μL/min, P < 0.05 and 6.5 ± 2.2 μL/min, P < 0.001, respectively). Calculated RA flow returned to baseline during hypercarbia/normoxia II and normoxia/normocarbia II. During hypercarbia/hyperoxia-300, calculated RA blood decreased significantly compared with normoxia/normocarbia II (P < 0.01) but not compared with normoxia/normocarbia I (Fig. 4) . 

Hypercarbia/normoxia I increased RA blood velocity by +17% ± 14% and calculated flow by +21% ± 19% from baseline. Hypercarbia/hyperoxia-500 reduced diameter by −8% ± 5%, velocity by −16% ± 17%, and calculated flow by −29% ± 14% from baseline values. There were strong correlations between P_ETO₂ and RA diameter, blood velocity, and calculated flow (r = −0.762, r = −0.675 and r = −0.807, respectively, P < 0.0001). 

MCA Vascular Reactivity

During hypercarbia/normoxia I, MCA blood velocity (MCAV) increased to 76 ±15 cm/s from baseline 54 ± 13 cm/s (P < 0.001) and then remained unchanged during hypercarbia/hyperoxia-500 and hypercarbia/normoxia II (82 ± 21 and 79 ± 19 cm/s, respectively, P < 0.001). During normocarbia/normoxia II, MCAV returned to baseline and then increased to 81 ± 19 cm/s during hypercarbia/hyperoxia-300 (P < 0.001; Fig. 5 ). 

The relative increases in MCAV were +45% ± 20% and +48% ± 18% during hypercarbia/normoxia I and II, respectively, and +54% ± 15% and +54% ± 20% during hypercarbia/hyperoxia-500 and hypercarbia/hyperoxia-300, respectively (Fig. 6) . Hypercarbia increased MCAV by 4.6% per mm Hg P_ETCO₂. The increase in MCAV to normoxic hypercarbia was greater than that of the RA (P < 0.001). There was strong correlation between P_ETCO₂ and MCAV (r = 0.964, P < 0.0001). 

Discussion

There were two major findings in the study. Hyperoxia reversed the hypercarbia-induced increase in RA blood flow in a concentration-dependent manner. Identical hyperoxic-hypercarbic stimuli caused profoundly different responses of MCA and RA blood velocities. 

First, we confirmed our previous findings that hyperoxia is a stronger vasoactive stimulus than hypercarbia in the retinal circulation and that with combined hyperoxia/hypercarbia, the constrictor effect of O₂ on RAs predominates over the vasodilator effect of CO₂. A limitation of our previous study was that the hyperoxic stimuli followed consecutively and progressively without a return to baseline. ²⁰ In the present study, we reversed the order of the application of the O₂ stimuli, giving the higher Po ₂ first and returning conditions to baseline between stimuli to minimize any persistent effects of arterial oxygen saturation on retinal and cerebral blood flow. Previously, we found that a 23% increase in P_ETCO₂ at a P_ETO₂ of 556 mm Hg caused a 36% decrease in retinal blood flow ²⁰ ; in this study, a 22% increase in P_ETCO₂ at a P_ETO₂ of 483 mm Hg decreased retinal blood flow by 29%. These results support the reproducibility of our retinal hemodynamic assessment technique. Of interest, we found less hypercarbia-induced increase in blood flow after exposure to the hypercarbia/hyperoxia-500 stimulus, suggesting a persistent vasoconstrictive effect of O₂ on retinal arterioles. The underlying mechanisms of cerebral and ocular vascular regulation in response to hyperoxia and hypercarbia are not completely understood. In both vascular beds, the primary involvement of endothelium-derived nitric oxide and endothelin (ET)-1 has been demonstrated. ^{19 25 26 27} Endothelin-1 is a potent vasoconstrictor with lasting effect, ^{28 29} whereas nitric oxide has a much shorter half life, ³⁰ which may explain why hyperoxia causes a concentration-dependent cumulative response in the retinal circulation. Moderate hyperoxia (P_ETO₂ 300 mm Hg) combined with hypercarbia induced mild vasoconstriction in the retina relative to the preceding normocarbic/normoxic II condition; however, there was no change in calculated RA blood flow between the hypercarbia/hyperoxia-300 stimulus and baseline. Of note, there was no difference in flow between normocarbia/normoxia I and normocarbia/normoxia II, suggesting that the retinal blood flow response to hypercarbia/hyperoxia-300 was not uninfluenced by possible persistent ET-1 effects. 

The second major finding is the discrepancy in responses of two organs of close embryologic origin, whose vascular beds are still widely assumed to behave congruently, even though conflicting evidence was available as early as 1964. ³¹ The present study is the first in which two identical provocative protocols were performed and the vascular reactivity of the RA and MCA compared in the same subjects. Both the superior-temporal RA and MCA velocities increased substantially in response to hypercarbia. However, hypercarbia induced greater increase in blood velocity in MCA than RA. The degree of MCA vascular reactivity agreed with results from previous studies. ^{32 33} Although adding hyperoxia to the hypercarbia constricted the RA, it had no discernible effect on MCA velocity. 

The MCA, one of the larger basilar vessels, receives 80% of the internal carotid artery blood flow. ³⁴ It functions as a conductive vessel (as opposed to a resistance vessel), and thus, without exception, changes in diameter are reported to be considerably less than 5% ^{35 36 37 38 39 40 41} under conditions that affect resistance vessels such as hypercapnia, ⁴⁰ hypocapnia, ^{38 40} hypertension, ³⁵ hypotension, ^{35 40} and pharmacologic vasodilator and vasoconstrictor ³⁵ provocations. As the MCA in adults is a rather large vessel (1.5–3 mm in diameter), such diameter changes would result in less than 4% to 6% change in MCAV reading for the same flow. The MCA diameter under these conditions has been monitored by ultrasound power analysis, ^{36 37} MRI scanning, ^{38 40} and even by direct observation under a microscope. ³⁵ Furthermore, in subjects undergoing provocations to alter cerebral blood flow, there is a strong correlation between resultant changes in MCAV and changes in cerebral blood flow as measured by “reference standard” techniques such as 133Xe SPECT ^{42 43} and electromagnetic flow probes on ipsilateral carotid artery ^{34 42} , with good correlations of r = 0.85, P < 0.001 ^{42 43} and r ² = 0.898. ^{34 44} Although historically debated, ⁴⁵ the overwhelming balance of experimental evidence leaves little doubt that in most conditions, changes in transcranial Doppler velocity signal are directly related to changes in cerebral blood flow. 

Retinal and cerebral vascular reactivity was assessed on two separate visits because simultaneous measurements of retinal and cerebral hemodynamics were ergonomically difficult. Nevertheless, the considerable similarity of achieved gas provocation parameters when making measurements at the two vascular sites minimizes the effect of this potential limitation. Another limitation of the present study is the use of a nonblinded observer. Even though the observer was not blinded to the composition of the inspired gas, she was blinded to the quantitative measurements of retinal blood velocity, thus diminishing the possible effect of observer bias. The study may also be more comprehensive if the retinal capillary, choroidal, and ophthalmic artery vascular reactivity could be assessed along with that of the retinal arterioles. Unfortunately the comfort level of subjects and the time needed to obtain good-quality recordings limited the number of vascular sites that could be examined. 

Part of the underlying motivation for this work was a desire to reveal the effect of carbogen (a mixture of approximately 1–5% CO₂ in O₂) on the vasculature when administered for therapeutic benefit. ²⁰ In addition to its direct action of vasoconstriction, O₂ causes hyperventilation and a reduction in P_ETCO₂ ⁴⁶ with a consequent additive vasoconstrictor effect of hypocapnia. ⁴⁷ This has prompted the suggestion to maintain isocapnia with O₂ administration. Moreover, since hypercarbia promotes vasodilation, ⁴⁸ CO₂ is often administered in conjunction with O₂ in the form of carbogen in an attempt to optimize tissue oxygenation. ^{49 50 51} First, Prisman et al. ⁵² have shown that carbogen does not reliably change P_ETCO₂ or arterial PCO₂ when administered to otherwise healthy subjects. Thus, vasodilatation cannot be presumed on the basis of carbogen administration without measuring P_ETCO₂ or arterial PCO₂. Second, we demonstrated that with respect to the eye, during combined hypercarbic/hyperoxic stimuli, as in the administration of carbogen, the vasoconstrictor effect of the O₂ component overpowers any hypercarbia-induced vasodilation. Thus, carbogen, while possibly increasing oxygenation and perfusion of cerebral tissue. ⁵³ may well have a deleterious effect on the retinas of patients with already compromised retinal arterioles, such as in atherosclerosis, vasculitis, or retinal vascular occlusive disease. Our results suggest that in healthy eyes using CO₂/O₂ mixtures that are only modestly hyperoxic (e.g., 40%) may achieve improved oxygenation but avoid retinal vasoconstriction. Conversely, the titrated retinal vasoconstriction in response to conventional carbogen mixtures may be desirable along with improved oxygenation in diseases characterized by retinal hyperperfusion, such as diabetic macular edema. ⁵⁴  

The reasons for the difference in vascular reactivity between the RA and MCA are yet to be determined. The RA lumen diameter is approximately 100 μm, and it functions as a resistance vessel, whereas the MCA has a lumen of approximately 1.5 to 3 mm, and it functions essentially as a conductance vessel with minimal change in diameter in response to changes in PCO₂. Nevertheless changes in the MCAV reflect the downstream hemodynamic responses in the cerebral microvasculature that exhibits similar reactivity to hypercarbia but insensitivity to hyperoxia relative to that of retinal arterioles. Furthermore, O₂ toxicity is an underappreciated risk factor for the retinal vascular bed. Prolonged hyperoxia causes concentration-dependent photoreceptor death in the adult rat retina. ⁵⁵ Hyperoxia-induced vasoconstriction in the retina protects against any intraretinal increase of Po ₂. The effectiveness of this mechanism probably fails at very high Po ₂ and in disease. In animal studies, inner retinal PO₂ is relatively unchanged during moderate hyperoxia (e.g., 40%), but is elevated at higher concentrations. ⁵⁶ In diabetic patients measurement of preretinal vitreous oxygenation during pure oxygen breathing revealed higher and progressively increasing concentrations compared with healthy subjects. ⁵⁷ Moreover, Arden et al. ¹ pointed out that high O₂ consumption by photoreceptors with relative intraretinal hypoxia is the main factor differentiating the retinal microcirculation from cerebral or other vascular beds. The active maintenance of these relative hypoxic conditions in the retina may explain the difference in vascular reactivity between the retinal and cerebral circulations. 

The results of the present study imply that, with respect to the eye, one must choose between increasing perfusion and increasing Po ₂. Hemoglobin is almost fully saturated at a PO₂ of approximately 100 mm Hg in patients without significant lung disease. As such, increasing arterial PCO₂ will increase perfusion and thus O₂ delivery. In the presence of lung disease, it is rational to increase the inspired PO₂ to maintain arterial PO₂ near 100 mm Hg only, to maintain the efficacy of hypercarbia on blood flow. Administering O₂ in an attempt to increase the volume of O₂ dissolved in the plasma also increases the PO₂, which will offset the small increase in blood O₂ content by a substantial reduction in retinal blood flow. Thus breathing pure O₂ may actually result in a net decrease O₂ delivery to the inner retina. ³¹ Now that simple methods of independently controlling PCO₂ and PO₂ are available, ⁵⁸ we may have a more effective alternative to carbogen for optimizing ocular blood flow. 

 Supported by Canadian Institutes of Health Research (CIHR) Operating Grant (CH) and CIHR New Investigator Award (CH), Canada Foundation for Innovation Infrastructure Award (CH), Vision Science Research Program Doctoral Scholarship (MK) and CIHR Strategic Training Program in Vision Health Research Fellowship (MK).

 Submitted for publication February 8, 2008; revised April 20, 2008; accepted September 19, 2008.

 Disclosure: M. Kisilevsky, None; A. Mardimae, None; M. Slessarev, P; J. Han, None; J. Fisher, P; C. Hudson, P

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Breathing protocol. A, normocarbia/normoxia I; B, hypercarbia/normoxia I; C, hypercarbia/hyperoxia-500; D, hypercarbia/normoxia II; E, normocarbia/normoxia II; F, hypercarbia/hyperoxia-300.

Table 1.

 

Effect of Inhalation of Air, Hypercarbic, and Combined Hypercarbic/Hyperoxic Gas Mixtures on Respiratory Parameters

	Normocarbia/Normoxia I	Hypercarbia/Normoxia I	Hypercarbia/Hyperoxia-500	Hypercarbia/Normoxia II	Normocarbia/Normoxia II	Hypercarbia/Hyperoxia-300
RA Vascular Reactivity Study
PETCO2 (mm Hg)	40.9 (0.7)	50.1 (0.9)*	49.9 (1)*	49.6 (1)*	41.2 (0.9)	49.7 (1)*
Increase in PETCO2 (%)	—	23 (3)	22 (3)	21 (3)	0.7 (1.8)	21 (3)
PETO2 (mm Hg)	104 (4)	104 (6)	483 (17)*	105 (4)	104 (6)	291 (9)*
RR (min)	17 (3)	20 (3), †	21 (3), †	22 (4), †	18 (3)	21 (3), †
MCA Vascular Reactivity Study
PETCO2 (mm Hg)	40.2 (0.8)	50.3 (1)*	49.8 (0.6)*	50.0 (0.6)*	40.3 (0.9)	50.2 (0.9)*
Increase in PETCO2 (%)	—	25 (4)	24 (2)	24 (3)	0.3 (3)	25 (3)
PETO2 (mm Hg)	102 (4)	106 (6)	475 (37)*	110 (7)	106 (7)	305 (6)*
RR (min)	16 (4)	18 (4)	20 (5), †	22 (7), †	18 (7)	20 (6), †

 Data are expressed as the mean (SD). Level of significance of change from baseline:

*   P < 0.0001;

†   P < 0.001.

Table 2.

 

Effect of Inhalation of Air, Hypercarbic, and Combined Hypercarbic/Hyperoxic Gas Mixtures on Systemic Hemodynamic Parameters

	Normocarbia/Normoxia I	Hypercarbia/Normoxia I	Hypercarbia/Hyperoxia-500	Hypercarbia/Normoxia II	Normocarbia/Normoxia II	Hypercarbia/Hyperoxia-300
RA Vascular Reactivity Study
SP (mm Hg)	117 (10)	125 (11)*	127 (10), †	127 (10), †	122 (11)	129 (13), †
DP (mm Hg)	69 (8)	72 (12)	75 (9)*	76 (9)*	74 (8)	79 (10), †
MAP (mm Hg)	87 (7)	91 (10)	92 (8)	94 (9)*	91 (9)	97 (9), †
HR (beats/min)	63 (9)	66 (10)	67 (10)	71 (8), †	69 (9), †	64 (8)
MCA Vascular Reactivity Study
SP (mm Hg)	125 (10)	128 (7)	132 (9)*	134 (11)*	131 (12)	133 (11)*
DP (mm Hg)	72 (6)	76 (6)	78 (6)*	79 (6)*	73 (5)	79 (7)*
MAP (mm Hg)	89 (5)	93 (5)	94 (5)	95 (7)*	90 (8)	95 (7)*
HR (beats/min)	74 (9)	72 (8)	70 (9)	73 (9)	72 (7)	75 (8)

 All data are expressed as the mean (SD). Level of significance of change from baseline:

*   P < 0.05;

†   P < 0.001. Data in bold, level of significance of change between the two study days: P < 0.05.

Figure 2.

 

RA diameter. A, normocarbia/normoxia I; B, hypercarbia/normoxia I; C, hypercarbia/hyperoxia-500; D, hypercarbia/normoxia II; E, normocarbia/normoxia II; F, hypercarbia/hyperoxia-300. *P < 0.001, **P < 0.01, ***P < 0.05.

Figure 3.

 

RA blood velocity. A, normocarbia/normoxia I; B, hypercarbia/normoxia I; C, hypercarbia/hyperoxia-500; D, hypercarbia/normoxia II; E, normocarbia/normoxia II; F, hypercarbia/hyperoxia-300. *P < 0.001, ***P < 0.05.

Figure 4.

 

RA blood flow. A, normocarbia/normoxia I; B, hypercarbia/normoxia I; C, hypercarbia/hyperoxia-500; D, hypercarbia/normoxia II; E, normocarbia/normoxia II; F, hypercarbia/hyperoxia-300. *P < 0.001, **P < 0.01, ***P < 0.05.

Figure 5.

 

MCA blood velocity. A, normocarbia/normoxia I; B, hypercarbia/normoxia I; C, hypercarbia/hyperoxia-500; D, hypercarbia/normoxia II; E, normocarbia/normoxia II; F, hypercarbia/hyperoxia-300. *P < 0.001.

Figure 6.

 

Relative responses of RA and MCA blood velocities. A, normocarbia/normoxia I; B, hypercarbia/normoxia I; C, hypercarbia/hyperoxia-500; D, hypercarbia/normoxia II; E, normocarbia/normoxia II; F, hypercarbia/hyperoxia-300.

The authors thank Steve Iscoe, Associate Professor of Physiology, Queen’s University, Kingston, Canada, for reviewing and suggesting modifications to a draft version of the manuscript.