点击显示 收起
Divisions of Pulmonary and Critical Care Medicine and Cardiology, Department of Medicine, University of Maryland, Baltimore, Maryland
ABSTRACT
Rationale and Objectives: Obstructive sleep apnea (OSA) is associated with oxidative stress and myocardial dysfunction. We hypothesized that the chronic intermittent hypoxia (CIH) component of OSA is sufficient to lead to these adverse effects.
Methods and Results: Rats were exposed to CIH (nadir O2, 4eC5%) for 8 hours/day, 5 days/week, for 5 weeks. Results were compared with similarly handled controls (HC). Outcomes included blood pressure (tail cuff plethysmograph), echocardiographic and invasive measures of left-ventricular (LV) function, and indices of oxidative stress that included levels of myocardial lipid peroxides and Cu/Zn superoxide dismutase. Blood pressure was greater in CIH (n = 22) than in HC (n = 22) after 2 weeks of exposure (136 ± 12 vs. 128 ± 8 mm Hg; p < 0.05). However, the difference disappeared by 5 weeks (127 ± 13 vs. 127 ± 13 mm Hg). LV weight/heart weight was greater with CIH (CIH, 0.52 ± 0.05; HC, 0.47 ± 0.06; p < 0.005). Echocardiograms revealed LV dilation, as well as decreased LV fractional shortening (CIH, 29.7 ± 9.8%; HC, 37.4 ± 7.1%; p < 0.001). LV end-diastolic pressure was increased with CIH (CIH, 13.7 ± 5.5; HC, 8.0 ± 2.9 mm Hg; p < 0.001), decreased LV dp/dtmax (CIH, 5072 ± 2191; HC, 6596 ± 720 mm Hg/second; p < 0.039), and decreased cardiac output (CIH, 48.2 ± 10.5; HC, 64.1 ± 10.9 ml/minute; p < 0.001). LV myocardial lipid peroxides were greater (CIH, 1,258 ± 703; HC 715 ± 240 e/mg protein; p < 0.05) and LV myocardial superoxide dismutase levels were lower (CIH, 10.3 ± 4.9; HC, 18.6 ± 8.2 U/mg protein; p < 0.05) with CIH.
Conclusions: CIH leads to oxidative stress and LV myocardial dysfunction.
Key Words: left ventricular function obstructive sleep apnea oxidative stress
Obstructive sleep apnea (OSA) is a common disorder (1) and is a risk factor for cardiovascular diseases, including hypertension, coronary artery disease, congestive heart failure, heart attacks, and stroke (2eC7). The three major components of OSA associated with cardiovascular consequences are large swings in intrathoracic pressure, postapneic arousals, and chronic intermittent hypoxia (CIH) (2). In animals, CIH has been shown to be associated with hypertension and sympathetic activation (8eC13), both of which could influence cardiac function. However, there are little published data on cardiac function in this model.
Oxidative stress, an important component of the generation of cardiovascular complications, such as atherogenesis, tissue damage, and ischemiaeCreperfusion injury (14), is defined as an increase in the cellular steady-state concentration of reactive oxygen species resulting from incomplete reduction of molecular O2 (15). Reactive oxygen species include the superoxide and hydroxyl radicals, hydrogen peroxide, and, after scavenging of the superoxide radical by nitric oxide, peroxynitrite. These reactive molecules are cytotoxic, causing lipid peroxidation, DNA damage and mutagenesis, depletion of intracellular ATP, alterations in calcium homeostasis, protein oxidation and apoptosis, and tissue necrosis. Reactive oxygen species are generated in the cell by several mechanisms, including the xanthine-hypoxanthine system (16), mitochondrial respiratory enzymes, lipoxygenase, certain P-450 enzymes, nitric oxide synthase, and membrane-bound reduced nicotinamide adenine dinucleotide phosphate oxidase (17, 18). Among the endogenous antioxidant systems, superoxide dismutase (SOD) detoxifies the superoxide radical, forming hydrogen peroxide, which is further metabolized by catalase to H2O.
Reactive oxygen species are produced in cell culture with short periods of hypoxia (19) and in neural tissue (20) with exposure to CIH. Furthermore, oxidative stress (21eC23) and decreased cellular antioxidant capacity (24) have been demonstrated in patients with OSA. However, clinical studies have not elucidated the stimuli responsible for oxidative stress due to the confounding influences of mechanical stresses, postapneic arousals, and associated medical conditions producing oxidative stress in humans with OSA. In addition, as oxidative stress markers were only determined from peripheral blood in clinical studies, the relationships between oxidative stress and dysfunction in the myocardium cannot be determined. We used a previously developed rat model (9eC13) to analyze myocardial function during exposure to CIH and its potential association with increased myocardial oxidative stress. The rat model allowed us to control for many of the confounding effects of factors associated with human OSA. We hypothesized that CIH leads to the following conditions: (1) myocardial dysfunction, (2) oxidative stress, and (3) decreased antioxidant capacity.
This study was presented, in part, in abstract form at the Spring 2005 meetings of the American Thoracic Society (25).
METHODS
CIH
Male Sprague-Dawley rats, 225eC275 g, were randomized into experimental (CIH) and similarly handled controls (HC) (12, 13, 26). During daylight hours, O2 concentration was decreased to 4eC6% approximately every 60 seconds (CIH) or was left unchanged (HC). Exposure was performed 8 hours/day, 5 days/week, for 5 weeks. Weekly, animals were weighed, and systolic arterial blood pressure (SBP) and heart rate measured (by tail cuff). At 5 weeks, animals were anesthetized for echocardiography and cardiac catheterization. M-mode echocardiographic measurements (see Figure 1 for example) included the following: end-diastolic (LVDd) and end-systolic (LVSd) left-ventricular (LV) anterioreCposterior cavitary dimension, LV fractional shortening ([LVDd eC LVSd]/LVDd x 100), end-diastolic (LVEDV) and end-systolic (LVESV) volumes ([7 x D3/(2.4 + D)] [27]), and LV ejection fraction ([LVEDV eC LVESV]/LVEDV x 100.
At catheterization, we measured blood pressure and LV pressures (micromanometer-tipped catheter), cardiac output (thermodilution), and maximum dp/dt of ventricular pressure rise (dp/dtmax). Stroke volume and systemic vascular resistance were calculated.
After the rats were killed, their hearts were removed and perfused via the ascending aorta with saline, the perfusate drained, and the hearts weighed. The separated LV and right ventricle (RV) were stored at eC70°. For analysis, heart tissues were minced and quickly homogenized using a microfuge tube pestle on ice.
As an index of myocardial tissue oxidative stress, total lipid peroxides (LPO) were measured as the sum of malonaldehyde and 4-hydroxyalkenals (28) using a commercially available kit (Bioxytech LPO-586; Oxisresearch, OXIS International, Inc., Portland, OR). Protein concentration was determined using a Protein Assay kit (Bio-Rad Laboratories, Hercules, CA) (29). Lipid peroxides were expressed as e/mg protein. As an index of antioxidant capacity, we measured cytosolic (Cu/Zn) SOD using a commercially available kit (Bioxytech Oxisresearch SOD-525) (30) after extraction with ethanoleCchloroform to inactivate mitochondrial and extracellular SOD.
Protocols
Hemodynamics, LV function, and oxidative stress with exposure to CIH.
Two groups of 22 animals each were exposed to CIH or HC, respectively, for 5 weeks. Subsets were used for measures of invasive and noninvasive cardiac function, SOD, and LPO (n indicated in RESULTS). Noninvasive measurements of blood pressure were validated by comparing simultaneous blood pressure measured via tail cuff with those measured via micromanometer-tipped catheter in three anesthetized rats while altering intravascular volume via saline infusion and phlebotomy.
To evaluate the development of LV dysfunction in two groups of 6 rats each (CIH or HC), we measured echocardiographic indices at baseline, after 3 weeks, and at 5 weeks of exposure.
Data were compiled and expressed as mean ± SD. Two-factor analysis of variance with post hoc analysis (Newman-Keuls) was used to analyze variables over time and Student's t test was used for variables measured at a single time point. Regression analysis was performed using the least-squares technique. The null hypothesis was rejected at the 5% level.
Studies were approved by the institutional animal care and use committee.
RESULTS
Table 1 shows the results of weekly measurements of bodyweight, SBP, and heart rate in the Group 1 animals. Two-way analysis of variance revealed that the overall difference between CIH and HC was significant (p < 0.001). Values at all time points were significantly different from their respective controls. However, the only single time point at which the difference between CIH and HC was significant was 2 weeks (p < 0.05). There were no changes in heart rate with time or between groups. Finally, body weight increased significantly in both groups, with no significant differences between the groups. Analysis of heart weight showed increased heart-toeCbody weight ratio, increased LVeCtoeCtotal heart weight ratio, and increased LV-to-RV weight ratio in the CIH-exposed compared with HC rats after 5-week exposure (Table 2).
Table 3 shows the 5-week catheterization data in the two groups. Arterial pressure and LV end-diastolic pressure were significantly increased in CIH compared with HC. Compared with HC, cardiac output and derived values and dp/dtmax were decreased with CIH. Echocardiographic studies showed that, compared with HC, CIH animals demonstrated increased LV cavitary volumes and decreased fractional shortening and ejection fraction (Table 4).
Whereas the CIH and HC groups showed similar levels of LV LPO (Figure 2) and LV SOD (Figure 3) after 1 day of exposure, LV LPO levels were significantly greater with CIH after 1 week and 5 weeks of exposure and LV SOD levels were lower with CIH after 5 weeks of exposure. A similar pattern was seen in RV LPO levels (not shown). In the RV, at 5 weeks, LPO was greater with CIH (881 ± 491 e/mg protein) than with HC group (573 ± 210 e/mg protein; p = 0.049). RV SOD levels tended to be lower after 5 weeks CIH (14.1 ± 7.1 units/mg protein) than for HC (19.1 ± 5.9 U/mg protein), but the difference was not significant (p = 0.15). The correlation between LV myocardial oxidative stress (LPO) and LV function (fractional shortening) at 5 weeks is shown in Figure 4. For LPO versus LV function, the regression was significant using a power function. As shown, greater LPO levels (oxidative stress) were associated with decreased myocardial function.
Figure 5 shows the time course of changes in echocardiographic measures (n = 6 for each group). For end-diastolic LV anterioreCposterior cavitary dimension, there was a significant increase over time for both CIH and HC (two-way analysis of variance). However, by post hoc analysis (Newman-Keuls), for HC, only the change at 3 weeks was significantly different from its baseline. At 5 weeks, the difference between CIH and HC was significant by post hoc analysis (Newman-Keuls, p < 0.05). There was a significant increase in end-diastolic LV anterioreCposterior cavitary dimension over time (p < 0.01). However, by post hoc analysis, for HC, the differences at 3 and 5 weeks were not significantly different from baseline. By post hoc analysis, the differences between HC and CIH were significant (p < 0.05) at 3 and 5 weeks. For LV fractional shortening and LVEF (derived from dimensions), there was a significant decrease over time (p < 0.01). Post hoc analysis revealed that only the difference between 5 weeks and baseline for CIH was significant. At 5 weeks, post hoc analysis showed that the difference between CIH and HC was significant (p < 0.05).
Figure 6 shows a comparison of systolic blood pressure measurements made simultaneously by tail cuff and by intraarterial catheter. The slope of the line was not significantly different than 1, showing that the tail cuff accurately reflected changes in blood pressure made using intraarterial catheter. However, there was an offset of 28.8 mm Hg, showing that the tail cuff systematically underestimated intraarterial pressure by approximately this amount.
DISCUSSION
We used a well-described rat model of intermittent hypoxia (9eC13) to study the mechanisms of myocardial dysfunction in OSA. Using this model, which isolates the intermittent hypoxia component of sleep apnea, we demonstrated that the development hypertension, and of myocardial dysfunction, is time-dependent and is associated with increased oxidative stress and decreased antioxidant activity. We also demonstrate a correlation between the degree of cardiac dysfunction and the degree of myocardial oxidative stress. However, as association does not prove causality, additional studies would be needed to confirm a causal relationship between oxidative stress and myocardial dysfunction. The ensuing discussion considers these findings in the light of the limitations of the study and the currently available literature.
We observed an increase in tail-cuff SBP in both HC and CIH animals (Table 1). However, early on, the change was greater with CIH. Although we do not know why SBP increased in HC, this has been observed before (31) and could be due to the stress of handling or the effects of aging and growth. However, the early (2-week) response with CIH is consistent with a greater degree of sympathoadrenal stimulation with CIH than with HC. The decrease toward baseline of SBP by 5 weeks likely reflects adaptive mechanisms, including baroreceptor adaptation, parasympathetic upregulation, and increased synthesis of nitric oxide at least in the rat model of CIH (31). However, for all other time points, there was no difference in noninvasively measured blood pressure (BP) between CIH and HC. To the extent that the noninvasive measurements of BP reflect the afterload placed on the LV, this would suggest that changes in cardiac function and oxidative stress at 5 weeks are not due to differences in LV afterload.
There were no heart-rate differences between CIH and HC. Although this might, at first, be considered inconsistent with heightened sympathoadrenal tone with CIH, in other studies using this model, heart rate does not change (9eC15), even though the role of sympathetic tone has been clearly documented. There may be adaptive mechanisms with chronic sympathetic stimulation that dampen the heart-rate response in rats.
We observed that, at 5 weeks, SBP measured at catheterization was greater with CIH than with HC. These data in anesthetized animals are similar to those reported by Greenberg and coworkers (12), but appear in contradiction to tail-cuff measurements of SPB. However, BP is determined by the balance of numerous factors, including autonomic tone, renal salt/water handling, and reflex inputs (e.g., baroreceptors). These vary from conscious to anesthetized state, and the balance of factors determining SBP is different with anesthesia than that while awake. It is possible that the higher SBP recorded under anesthesia with CIH compared with HC represents unmasking of differences in sympathoadrenal tone engendered by CIH, a difference obscured by increased activity of other homeostatic mechanisms while awake.
LV function was adversely affected by CIH compared with HC, as indicated by a number of invasively and noninvasively measured indices. Hypertrophy was demonstrated by increased LV/heart weight (Table 2), dilation by increased diastolic and systolic chamber size (Table 4), and decreased function by increased LV end-diastolic pressure, decreased dp/dtmax, and cardiac output (Table 3). What are the possible causes of diminished LV function with CIH We do not believe that differences in LV afterload (BP) could account for the observed changes. We believe that the noninvasive measurements of BP more accurately reflect LV afterload during the exposure period, as these are measured in the nonanesthetized state. Although there were differences in noninvasive BP at 2 weeks, there were none at any other time period. Although we cannot totally rule out some residual effect of early BP differences, it seems unlikely that these could cause changes in LV function of the magnitude we observed at 5 weeks (e.g., 22% decrease in fractional shortening and a 50% increase in LV end-systolic volume). As noted above, during anesthesia (invasive measures), BP was observed to be higher in CIH than in HC. However, the magnitude of this difference in mean arterial pressure was small (10 mm Hg; 10%). It is not likely that this small difference in LV afterload could have caused the large differences in invasive indices of LV function that we observed: 71% increase in LV end-diastolic pressure, 33% decrease in cardiac output, and 30% decrease in dp/dtmax. Hence, it seems more likely that the effects of hypoxia mediated changes in myocardial function.
In patients with OSA, myocardial dysfunction and hypertrophy have been well documented in many studies and attributed to the effects of hypertension, mechanical swings in intrathoracic pressure, and sympathetic stimulation (reviewed in References 2, 32, and 33). Our study demonstrates that sequalae of the intermittent hypoxia of OSA are sufficient to lead to myocardial dysfunction, although other factors, such as obesity and concomitant medical conditions, may certainly contribute as well in clinical disease.
The mechanisms by which CIH produces oxidative stress are not well characterized. Certainly, intermittent hypoxia resembles the classic model of myocardial ischemiaeCreperfusion injury (see References 34eC36 for recent reviews). Much attention has been focused on the role of reperfusion in generation of reactive oxygen species. In the setting of ischemia and reperfusion, the carefully coordinated homeostasis between reactive oxygen species and endogenous antioxidants is disturbed. The reintroduction of molecular oxygen after ischemia is capable of producing excess reactive oxygen species that become capable of damaging cell components, such as lipids, DNA, and proteins, as well as initiating apoptosis. Recently, several investigators have observed reactive oxygen species generation during ischemia (37). Although, at first, this appears paradoxic, with ischemia, respiratory cytochromes become redox-reduced, allowing them to directly transfer electrons to oxygen (38). Considerable molecular oxygen would still be available even with ischemia, and likely more so with hypoxia. In the presence of molecular oxygen, a redox-reduced cell appears capable of producing large amounts of reactive oxygen species (38). Thus, both the hypoxia per se, and/or the reoxygenation associated with CIH could be capable of producing reactive oxygen species. On the other hand, because blood flow is stopped in the classic ischemiaeCreperfusion model, and not in the intermittent hypoxia model, there may be limits to which one can generalize from ischemiaeCperfusion to CIH as continued perfusion could allow for wash-out of potentially toxic metabolites.
Another mechanism for the production of reactive oxygen species with CIH is the increase in sympathoadrenal tone associated with CIH (8eC13). -Receptor stimulation can stimulate myocardial oxidative stress (39). During the acute phase of sympathetic stimulation, reactive oxygen species are important activators of second messengers (mitogen-activated protein kinase cascades) and, during the chronic phase, reactive oxygen species appear to participate in events leading to cardiac remodeling and fibrogenesis (39). Whatever the mechanisms producing oxidative stress with CIH, they appear operative in the two organs most affected by OSA, the heart (the present study) and the brain (20), suggesting that oxidative stress is a generalized phenomenon of CIH.
The correlation between tissue levels of LPO and LV function (Figure 4) is consistent with the notion that oxidative stress is a contributor to myocardial dysfunction with CIH. However, as association does not prove causality, additional studies would be needed to confirm a causal relationship between oxidative stress and myocardial dysfunction. We cannot rule out the possibility that both LV dysfunction and oxidative stress are due to a third factor, or that LV dysfunction may contribute to oxidative stress.
We found decreased activity of myocardial Cu/Zn (cytosolic) SOD with CIH. This finding is consistent with human studies demonstrating decreased antioxidant potential in patients with OSA (40). Indeed, a decrease in antioxidant potential could be an important contributor to oxidative stress in CIH. Although this could represent decreased gene expression, and/or increased breakdown, caution is urged before concluding that decreased SOD activity represents decreased antioxidant capacity. Antioxidant mechanisms are redundant, and studies should be performed assessing the effects of CIH on other endogenous and exogenous antioxidant systems. The finding that antioxidant treatment of rats exposed to CIH alleviated neurocognitive deficits (20) raises the possibility that antioxidants could be a useful adjunct treatment for patients with OSA.
The combined data of Figures 2, 3, and 5 show progressive deterioration of LV function preceded by biochemical alterations (oxidative stress, decreased SOD). Thus, it appears that oxidative stress begins earlier than echocardiographic signs of LV dysfunction, possibly due to increased expression of enzymes producing reactive oxygen species and/or downregulation of expression of antioxidant enzymes.
Critique of Methods
This study indicates that the consequences of CIH led to oxidative stress and LV dysfunction. Aside from the exposure to intermittent hypoxia, the groups were placed in similar chambers and exposed to the same conditions of ambient temperature, humidity, and noise (see METHODS in online supplement for details). Thus, we cannot attribute the findings to any other differences in handling except for the exposure to CIH versus controls.
We did not find evidence of RV hypertrophy as indicated by RV weight. Others (41) have reported increased pulmonary arterial pressure and RV hypertrophy with CIH, although this is not found in all studies (41). The rat RV is small and thin-walled, with a thickness close to the technical limits of our ability to record on echocardiogram. Hence, we were not able to accurately assess RV function echocardiographically and right-sided pressures and flows were not measured. We note that the trends in LPO and SOD were in the same direction as for the LV, suggesting that the same biochemical effects seen in the LV myocardium were also seen in the RV.
Figure 6 shows that, when measured simultaneously, SBP at catheterization was greater than that measured by tail cuff (both measurements under anesthesia). This likely reflects intrinsic differences in the measurement technique. The micromanometer-tipped catheter has a higher frequency response than the tail cuff plethysmograph, will record intraarterial reflected waves, and records end-pressures rather than side-pressures. These effects will produce a greater SBP recorded with the micromanometer-tipped catheter than that with the tail-cuff system. However, the change in simultaneously measured intraarterial and tail-cuff pressures was the same as blood volume was changed, lending credence to the findings over time using the noninvasive tail-cuff pressure technique.
Conclusions
This is the first demonstration that intermittent hypoxia, a feature of human OSA, leads to myocardial dysfunction, oxidative stress, and decreased activity of at least one major antioxidant system. Measures of oxidative stress are correlated with the degree of impairment of LV function. Further studies should be directed at the mechanisms of oxidative stress in CIH, the possible causal relationship between oxidative stress and LV dysfunction, and the potential benefits of antioxidant therapy in CIH and OSA.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournal.org
REFERENCES
Young T, Peppard PE. Epidemiological evidence for and association of sleep disordered breathing with hypertension and cardiovascular disease. In: Bradley TD, Floras JS, editors. Sleep apnea: implications in cardiovascular and cerebrovascular disease. New York: Marcel Dekker; 2000. pp. 261eC284.
Podszus TE, Greenberg H, Scharf SM. Influence of sleep state and sleep-disordered breathing on cardiovascular function. In: Sullivan CE, Saunders NA, editors. Sleep and breathing II. New York: Marcel Dekker; 1993. pp. 257eC310.
Lavie P, Here P, Hoffstein V. Obstructive sleep apnoea syndrome as a risk factor for hypertension: population study. BMJ 2000;320:479eC482.
Grote L, Hedner J, Peter JH. Sleep-related breathing disorder is an independent factor for uncontrolled hypertension. J Hypertens 2000;18:679eC685.
Shahar E, Whitney CW, Redline S, Lee ET, Newman AB, Javier Nieto F, O'Connor GT, Boland LL, Schwartz JE, Samet JM, for the Sleep Heart Health Study Research group. Sleep-disordered breathing and cardiovascular disease cross-sectional results of the Sleep Heart Health Study. Am J Respir Crit Care Med 2001;163:19eC25.
Moore T, Rabben T, Wiklund U, Franklin KA, Eriksson P. Sleep-disordered breathing in men with coronary artery disease. Chest 1996;109:659eC663.
Lavie L. Sleep apnea syndrome, endothelial dysfunction, and cardiovascular morbidity. Sleep 2004;27:1053eC1055.
Chen L, Sica AL, Scharf SM. Mechanisms of acute cardiovascular response to periodic apneas in sedated pigs. J Appl Physiol 1999;86:1236eC1246.
Fletcher EC, Lesske J, Qian W, Miller CC III, Unger T. Repetitive episodic hypoxia causes diurnal elevations of systemic blood pressure in rats. Hypertension 1992;19:555eC561.
Fletcher EC, Lesske J, Behm R, Miller CC III, Strauss H, Unger T. Carotid chemoreceptors, systemic blood pressure, and chronic episodic hypoxia mimicking sleep apnea. J Appl Physiol 1992;72:1978eC1984.
Fletcher EC, Lesske J, Culman J, Miller CC III, Unger T. Sympathetic denervation blocks blood pressure elevation in episodic hypoxia. Hypertension 1992;20:612eC619.
Greenberg HE, Sica A, Batson D, Scharf SM. Chronic intermittent hypoxia increases sympathetic responsiveness to hypoxia and hypercapnia. J Appl Physiol 1999;86:298eC305.
Sica AL, Greenberg HE, Ruggiero DA, Scharf SM. Chronic intermittent hypoxia: a model of sympathetic activation in the rat. Respir Physiol 2000;121:173eC184.
Griendling KK, FitzGerald GA. Oxidative stress and cardiovascular injury Part I: basic mechanisms and in vivo monitoring of ROS. Circulation 2003;108:1912eC1916.
Richter C, Gogvadze V, Laffranchi R, Schlapbach R, Schweizer M, Sufer M, Walter P, Yaffee M. Oxidants in mitochondria: from physiology to diseases. Biochim Biophys Acta 1995;1271:67eC74.
McCord JM. Oxygen-derived free radicals in postischemic tissue injury. N Engl J Med 1985;312:159eC163.
Ronson RS, Nakamura M, Vinten-Johansen J. The cardiovascular effects and implication of peroxynitrite. Cardiovasc Res 2000;44:47eC59.
Griendling KK, Sorescu D, Ushio-Fukai M. NAP(P)H oxidase: role in cardiovascular biology and disease. Circ Res 2000;86:494eC501.
Prabhaker NR, Kumar GK. Oxidative stress in the systemic and cellular responses to intermittent hypoxia. Biol Chem 2004;385:217eC221.
Row BW, Liu R, Xu W, Kneirandish L, Gozal D. Intermittent hypoxia is associated with oxidative stress and spatial learning deficits in the rat. Am J Respir Crit Care Med 2003;167:1548eC1553.
Lavie L. Obstructive sleep apnoea syndrome: an oxidative stress disorder. Sleep Med Rev 2003;7:35eC51.
Schulz R, Mahmoudi S, Hattar K, Sibelius U, Olschewski H, Mayer K, Seeger W, Grimminger F. Enhanced release of superoxide from polymorphonuclear neutrophils in obstructive sleep apnea. Am J Respir Crit Care Med 2000;162:566eC570.
Barcelo A, Miralles C, Barbe F, et al. Abnormal lipid peroxidation in patients with sleep apnoea. Eur Respir J 2000;16:644eC647.
Christou IK, Moulas AN, Pastaka C, Gouroulianis KI. Antioxidant capacity in obstructive sleep apnea patients. Sleep Med 2003;4:225eC228.
Scharf SM, Chen L, Einbinder E, Hasday J, Badke CW. Oxidative stress and left ventricular function in chronic intermittent hypoxia . Proc Am Thorac Soc 2005;2:A879.
Sica AL, Greenberg HE, Scharf SM, Ruggiero DA. Immediate-early gene expression in cerebral cortex following exposure to chronic-intermittent hypoxia. Brain Res 2000;870:204eC210.
Teichholz LE, Kreulen T, Herman MV, Gorlin R. Problems in echocardiographic volume determinations: echocardiographiceCangiographic correlations in the presence and absence of asynergy. Am J Cardiol 1976;37:7eC11.
Hoyos M, Guerrero JM, Perez-Cano R, Olivan J, Fabiani F, Garcia-Perganeda A, Osuna C. Serum cholesterol and lipid peroxidation are decreased by melatonin in diet-induced hypercholesterolemic rats. J Pineal Res 2000;28:150eC155.
Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of proteineCdye binding. Anal Biochem 1976;72:248eC254.
Okatani Y, Wakatsaki A, Kaneda C. Melatonin increases activities of glutathione peroxidase and superoxide dismutase in fetal rat brain. J Pineal Res 2000;28:89eC96.
Benoff B, Greenberg H, Sica A, Scharf S. Chronic intermittent hypoxia (CIH) in rats blunts the pressor response to NO synthase (NOS) inhibition . Am J Respir Crit Care Med 2000;161:A629.
Scharf SM. Effects of continuous positive airway pressure in animal models of congestive heart failure. In: Bradley TD, Floras JS, editors. Sleep apnea: implications in cardiovascular and cerebrovascular disease. New York: Marcel Dekker; 2000. pp. 435eC460.
Bradley TD, Floras JS. Pathophysiological interactions between sleep apnea and congestive heart failure. In: Bradley TD, Floras JS. Sleep apnea: implications in cardiovascular and cerebrovascular disease. New York: Marcel Dekker; 2000. pp. 385eC414.
Becker LB. New concepts in reactive oxygen species and cardiovascular reperfusion physiology. Cardiovasc Res 2004;61:461eC470.
Dhalla NS, Elmoselhi AB, Hata T, Makino N. Status of myocardial antioxidants in ischemiaeCreperfusion injury. Cardiovasc Res 2000;47:446eC456.
Chi NC, Karliner JS. Molecular determinants of responses to myocardial ischemia/reperfusion injury: focus on hypoxia-inducible and heat shock factors. Cardiovasc Res 2004;61:437eC447.
Kevin LG, Camara AK, Riess ML, Novalija E, Stowe DF. Ischemic preconditioning alters real-time measure of O2 radicals in intact hearts with ischemia and reperfusion. Am J Physiol Heart Circ Physiol 2003;284:H566eCH574.
Nohl H, Jordan W. The mitochondrial site of superoxide formation. Biochem Biophys Res Commun 1986;138:533eC539.
Zhang GX, Kimura S, Nishiyam A, Shokoji T, Rahman M, Yao L, Nagai Y, Fujisawa Y, Miyatake A, Abe Y. Cardiac oxidative stress in acute and chronic isoproterenol-infused rats. Cardiovasc Res 2005;65:230eC238.
Christou IK, Moulas AN, Pastaka C, Gouroulianis KI. Antioxidant capacity in obstructive sleep apnea patients. Sleep Med 2003;4:225eC228.
McGuire M, Bradford A. Chronic intermittent hypoxia increases haematocrit and causes right ventricular hypertrophy in the rat. Respir Physiol 1999;117:53eC58.