Portal de Periódicos da CAPES

OSA and cardiorespiratory fitness: a review

2021; American Academy of Sleep Medicine; Volume: 18; Issue: 1 Linguagem: Inglês

10.5664/jcsm.9628

1550-9397

Tyler Powell, Vincent Mysliwiec, Matthew Brock, Michael J. Morris,

Chronic Obstructive Pulmonary Disease (COPD) Research

Free AccessReview ArticlesOSA and cardiorespiratory fitness: a review Tyler A. Powell, MD, Vincent Mysliwiec, MD, Matthew S. Brock, MD, Michael J. Morris, MD Tyler A. Powell, MD Address correspondence to: Tyler A. Powell, MD, Sleep Medicine Service, Wilford Hall Ambulatory Service Center, JBSA Lackland Air Force Base, TX 78234; Email: E-mail Address: [email protected], E-mail Address: [email protected] Sleep Medicine Service, Wilford Hall Ambulatory Surgery Center, JBSA Lackland Air Force Base, Texas , Vincent Mysliwiec, MD Department of Psychiatry and Behavioral Sciences, University of Texas Health Science Center at San Antonio, San Antonio, Texas , Matthew S. Brock, MD Sleep Medicine Service, Wilford Hall Ambulatory Surgery Center, JBSA Lackland Air Force Base, Texas , Michael J. Morris, MD Graduate Medical Education, Brooke Army Medical Center, JBSA Fort Sam Houston, Texas Published Online:January 1, 2022https://doi.org/10.5664/jcsm.9628SectionsAbstractEpubPDF ShareShare onFacebookTwitterLinkedInRedditEmail ToolsAdd to favoritesDownload CitationsTrack Citations AboutABSTRACTThe effects of untreated obstructive sleep apnea (OSA) on cardiopulmonary function remain unclear. Cardiorespiratory fitness (CRF), commonly reflected by VO2 max measured during cardiopulmonary exercise testing, has gained popularity in evaluating numerous cardiopulmonary conditions and may provide a novel means of identifying OSA patients with the most clinically significant disease. This emerging testing modality provides simultaneous assessment of respiratory and cardiovascular function with results helping uncover evidence of evolving pathology in either organ system. In this review, we highlight the current state of the literature in regard to OSA and CRF with a specific focus on changes in cardiovascular function that have been previously noted. While OSA does not appear to limit respiratory function during exercise, studies seem to suggest an abnormal cardiovascular exercise response in this population including decreased cardiac output, a blunted heart rate response (ie, chronotropic incompetence), and exaggerated blood pressure response. Surprisingly, despite these observed changes in the cardiovascular response to exercise, results involving VO2 max in OSA remain inconclusive. This is reflected by VO2 max studies involving middle-aged OSA patients showing both normal and reduced CRF. As prior studies have not extensively characterized oxygen desaturation burden, we propose that reductions in VO2 max may exist in OSA patients with only the most significant disease (as reflected by nocturnal hypoxia). Further characterizing this relationship remains important as some research suggests that positive airway pressure therapy or aerobic exercise may improve CRF in patients with OSA. In conclusion, while it likely that severe OSA, via an abnormal cardiovascular response to exercise, is associated with decreased CRF, further study is clearly warranted to include determining if OSA with decreased CRF is associated with increased morbidity or mortality.Citation:Powell TA, Mysliwiec V, Brock MS, Morris MJ. OSA and cardiorespiratory fitness: a review. J Clin Sleep Med. 2022;18(1):279–288.INTRODUCTIONCardiorespiratory fitness (CRF), typically defined by the efficiency by which the cardiovascular and respiratory systems supply oxygen to skeletal muscle during exercise, is an emerging topic in preventive medicine. In a joint statement in 2016 by the European Association for Cardiovascular Prevention and Rehabilitation and the American Heart Association, cardiopulmonary exercise testing (CPET) to assess CRF was emphasized as a valuable tool in the management of numerous cardiopulmonary conditions (eg, chronic obstructive pulmonary disease, congestive heart failure, obesity).1 CRF, most commonly reflected by measurement of the maximum oxygen uptake (VO2 max) during CPET, is a reliable predictor of cardiovascular disease and mortality (perhaps even more so than traditional risk factors) and has shown promise in helping determine optimal timing for transplantation in heart failure.2–7 Given this rising interest in CRF as a prognostic tool, including in healthy patients, novel methods for determining VO2 max as well as demographic-specific normative values are being developed to aid interpretation.8–11 With these advances, routine assessment of CRF may become a reality, allowing for early recognition of heightened cardiovascular risk and implementation of preventive measures to reduce long-term morbidity and mortality. One condition where this may be particularly useful is obstructive sleep apnea (OSA), where CRF assessment may help identify patients at elevated cardiovascular risk whom positive airway pressure (PAP) or alternative OSA therapies may benefit most.OSA, characterized by repetitive partial or complete obstruction of the upper airway during sleep, impacts between 9% and 38% of the population and may elevate cardiovascular risk.12,13 Repetitive hypoxia-reoxygenation events and nocturnal arousals associated with OSA elevate serum inflammatory biomarkers, sympathetic nervous system activity, and endothelial dysfunction that together may contribute to atherogenesis.14–23 Literature evaluating the relationship between OSA and the development of various cardiovascular conditions is robust, with studies demonstrating an association between OSA and hypertension, atrial fibrillation, sudden cardiac death, and myocardial infarction.24–30 The extent to which this relationship is affected by OSA severity (defined by variables such as daytime sleepiness, apnea-hypopnea index [AHI], and degree of oxygen desaturation) remains debated, though it is likely that more severe disease portends greater cardiovascular risk.31,32 This association between OSA and clinical cardiovascular disease has stimulated interest in the effects of OSA on CRF as a potential prognostic and therapeutic tool. In the course of this review, we seek to present the current state of the literature in regard to OSA and CRF as defined by physiologic performance during CPET. While this discussion will primarily focus on VO2 max as a direct reflection of CRF, we will briefly explore prior findings involving indirect measures of CRF including heart rate and blood pressure response to peak exercise and respiratory efficiency parameters also collected during CPET. To aid in comparison, we will describe prior OSA study groups by both AHI and SpO2 nadir, two of the commonly used variables to describe OSA severity in prior research on this topic.RESULTSExercise and OSADuring CPET, graded exercise is performed with the use of a treadmill or cycle ergometer while respiratory gas exchange is measured via a nonrebreathing valve and metabolic cart.33 This setup allows for measurement of airflow volume as well as concentration of oxygen and carbon dioxide in expired gas which can be utilized to determine minute ventilation (VE), consumption of oxygen (VO2) and production of carbon dioxide (VCO2), respiratory exchange ratio (VCO2/VO2), and ventilation/carbon dioxide production ratio (VE/VCO2). These variables, along with continuous pulse oximetry monitoring (SpO2), provide information about exercise performance. Of these variables, VO2 provides the most comprehensive measure of cardiopulmonary performance during exercise and is represented by the following formula in which arterial and venous oxygen content are represented by CaO2 and CvO2:VO2=(HR×SV)×(CaO2−CvO2).During exercise, predictable increases in cardiac stroke volume, heart rate, tissue blood oxygen extraction, and VE occur until the maximum consumption of oxygen or VO2 max is achieved. Age, sex, and race all influence normative values for VO2 max, which may not be achieved by certain individuals due to cardiovascular, pulmonary, and metabolic disorders or due to physical deconditioning or submaximal patient effort.34 Suggested algorithms for CPET interpretation begin with an assessment of VO2 max, which if reduced (< 85% predicted) should be followed by a determination of the etiology of limitation. A gas exchange or pulmonary limitation is indicated when there is a decrease in SpO2 of greater than 4% (or to an absolute value less than 90%) at peak exercise or there is an elevation in peak VE to greater than 80% of maximum ventilatory volume, while a VE/VCO2 > 40 at peak exercise may suggest cardiovascular limitation. Nonlinear trends in exercise heart rate and blood pressure may also suggest cardiovascular limitation, while a peak heart rate less than 85% predicted maximum or a respiratory exchange ratio < 1.0 likely reflects submaximal exercise effort.1The mechanism by which OSA may limit exercise performance remains unclear. Interestingly, despite being a primarily pulmonary condition, OSA does not appear to limit the respiratory response to exercise. Studies of ventilatory response and gas exchange during peak exercise in patients with OSA have been predominantly normal, including those involving patients with the most severe disease (as represented by AHI and degree of oxygen desaturation). In a recent study by Han et al comparing patients with obesity hypoventilation syndrome (AHI 40.4 events/h, SpO2 nadir 71.9%) to those with eucapnic OSA (AHI 45.6 events/h, SpO2 nadir 72.4%) and obesity without OSA (AHI 3.9 events/h, SpO2 nadir 85.0%), the respiratory response to exercise was noted to be similar between groups as represented by peak VE/maximum ventilatory volume (57.2% vs 54.3% vs 49.6%, P > .05) and SpO2 at peak exercise (98.8% vs 98.5% vs 98.0%).35 Similarly, in a study by Bernhardt et al comparing 8 obese patients with OSA (AHI 25.4 events/h, SpO2 nadir not reported [NR]) to 8 age- and weight-matched controls (AHI 1.3 events/h, SpO2 nadir NR), peak VE (31.8 vs 26.2 L/min; P = .129) and tidal volume (1,277 vs 1,421; P = .947) were similar between groups.36Abnormal muscle metabolism has been postulated as a contributor to reduced exercise performance in OSA, though the research to support this hypothesis is limited. In a study involving needle biopsy of the quadriceps femoris of 12 consecutive patients with severe OSA (AHI 70 events/h, SpO2 nadir NR) and six healthy controls, the diameter of type II skeletal muscle fibers was smaller and activity of muscle metabolic enzymes comparatively less in patients with OSA.37 Likewise, in a study of 11 patients with moderate OSA (AHI 25.6 events/h, SpO2 nadir 85%), blood lactate concentration and rate of lactate elimination during exercise were significantly decreased in patients with OSA compared to controls, a finding which may suggest impaired glycolytic and oxidative metabolism in the skeletal muscle of patients with OSA.38 Studies involving normal patients without OSA exposed to hypoxic environmental conditions have also shown abnormalities in skeletal muscle function. In a study of high-altitude climbers with prolonged exposure to hypoxic conditions, muscle biopsies revealed a reduction in skeletal muscle fiber size and density as well as muscle oxidative capacity, changes that may occur with the intermittent hypoxia characteristic of moderate to severe OSA.39 These studies suggest that muscle structure and function may be altered in OSA; however, the potential of this mechanism to affect CRF in OSA remains unknown.The most convincing evidence to date supports a potential cardiovascular limitation to exercise in OSA. In a comprehensive study by Alonso-Fernández et al involving CPET with cardiac output measurement (by the CO2 rebreathing equilibrium method), it was found that nonhypertensive patients with untreated severe OSA (AHI 43.6 events/h, SpO2 nadir 72%) and normal resting left ventricular function had significant reductions in cardiac output at peak exercise compared to healthy controls.40 Interestingly, after the patients with OSA received 3 months of treatment with PAP, their left ventricular systolic performance during exercise significantly improved. Other studies have suggested subclinical abnormalities in resting left and right ventricular mass and systolic function in OSA, although these results have been inconsistent.41,42 In one study by Shivalkar et al, subclinical differences in stroke volume, intraventricular septum thickness, and right ventricular free wall motion were noted in patients with severe OSA (AHI 42 events/h, SpO2 nadir 76%) compared to controls, and these abnormalities improved with continuous PAP (CPAP) therapy.43 Decrements in left ventricular diastolic filling have also been previously noted in patients with moderate to severe OSA, with one study showing grade 1 diastolic dysfunction in more than 20% of these patients.44,45 In addition to these potential abnormalities in ventricular size and function, subclinical abnormalities in the pulmonary vasculature may also be present in untreated OSA, with one study showing a correlation between pulmonary arterial stiffness and OSA severity.46Multiple studies also suggest a blunted heart rate response, or chronotropic incompetence, to peak exercise in OSA, which may also be an indicator of cardiovascular limitation in this population. This finding is thought to be due to functional downregulation of cardiac β1-receptor activity from chronic sympathetic hyperactivity.47,48 In a study of middle-aged women with and without OSA undergoing exercise testing, peak heart rate was significantly lower in those with mild and moderate-to-severe OSA (148 and 144 beats per minute [bpm], respectively) compared to controls (158 bpm).49 Similarly, in a study comparing 21 patients with moderate-to-severe OSA (AHI 55 events/h, SpO2 nadir NR) to 10 healthy controls, chronotropic reserve (a marker of heart rate response to exercise) was significantly lower in patients with OSA (79.0% vs 99.0%; P = .01).50 Most convincingly, in a meta-analysis by Mendelson et al that evaluated patients with OSA with mild, moderate, and severe disease, it was found that patients with OSA, regardless of severity, had a significantly lower peak heart rate compared to controls with a mean difference of approximately 8 bpm (P = .02).51 This postulated downregulation of cardiac β1-receptor activity may not only limit peak exercise performance but also delay heart rate recovery (HRR) postexercise.52 Cholidou et al identified that patients with mild/moderate and severe OSA had progressively less HRR at 1, 2, and 3 minutes after peak exercise compared to controls without OSA (HRR at 1 minute: 29 [control] vs 24 [mild/moderate OSA] vs 20 [severe OSA] bpm; HRR at 2 minutes: 40 vs 34 vs 28 bpm; HRR at 3 minutes: 46 vs 39 vs 33 bpm; P < .05).53Studies showing an exaggerated blood pressure response to exercise in OSA may also support this hypothesis. Kasiakogias et al studied 57 patients with hypertension and moderate-to-severe OSA (AHI 30 events/h, SpO2 nadir 80.4%) and 57 hypertensive controls. They found peak systolic blood pressure during exercise was higher in patients with OSA (197.6 vs 187.8 mm Hg; P = .03), with significantly more patients with OSA having a hypertensive response to exercise, defined as peak systolic blood pressure ≥ 210 mm Hg (44% vs 19%; P = .009).54 Multivariate analysis in this study revealed that a hypertensive response to exercise independently correlated with AHI and SpO2 nadir. Similar results were noted in a CPET study by Przybyłowski et al involving 111 patients with OSA (AHI 47.2 events/h, SpO2 nadir 68.5%) in which a hypertensive response to peak exercise was noted in 35%.55 An abnormal diastolic blood pressure response in patients with OSA has similarly been noted. In a study of 17 normotensive patients with OSA (AHI 33.3 events/h, SpO2 nadir 89.6%) and 10 blood pressure-matched controls, a higher diastolic blood pressure at peak exercise was noted in patients with OSA (115 vs 101 mm Hg; P < .01), with patients with OSA reaching a diastolic blood pressure of 110 mm Hg at a lower VO2.56 This exaggerated diastolic blood pressure response to exercise was also noted in other studies by Vanhecke et al and Barros de Carvalho et al, though in the latter this was partially attributed to higher resting diastolic blood pressure.49,57VO2 max and OSADespite the aforementioned evidence linking untreated OSA with potential cardiovascular and muscular dysfunction during exercise (variables expected to limit VO2 max), studies involving VO2 max in patients with OSA have had variable results. An explanation for this variability remains unclear, though it may reflect differences in study demographics, OSA severity as reflected by degree of oxygen desaturation, and OSA phenotypic subtypes as previously suggested by Zinchuk and Yaggi.58 From a demographic standpoint, body weight, age, and sex have all been postulated to influence the cardiovascular effects of OSA, though the degree to which these variables influence the relationship between untreated OSA and VO2 max is not clearly defined. For body weight, this is made evident by the discordant results in prior studies comparing both lean and obese patients with OSA. Regarding age and sex, studies involving middle-aged male patients have predominantly shown a reduction in VO2 max, while more age- and sex-diverse studies have mostly not shown a reduction. Perhaps it is those patients with the more traditional definition of OSA (middle-aged obese males) who have the most significant effect on VO2 max.In regard to OSA severity, prior research on this topic has focused predominantly on AHI, which may limit comparability between studies. It is becoming apparent that OSA severity is more nuanced than previously believed, with continued evidence supporting that it is likely the degree of nocturnal hypoxia expressed by oxygen desaturation depth and duration, rather than the frequency of OSA events (as reflected by the AHI), that most significantly affects the cardiovascular system.59 Measures of hypoxia during polysomnography including SpO2 nadir, time spent with oxygen saturation < 90%, and oxygen desaturation index have all shown better predictability of cardiovascular outcomes in prior research compared to AHI, though thresholds of significance remain unclear.60 More recently, the "hypoxic burden" (a measure to quantify the depth and duration of nocturnal oxygen desaturation) has garnered attention as a reliable prognosticator of cardiovascular risk in OSA.61 Unfortunately, prior studies evaluating VO2 max in OSA have not extensively evaluated these hypoxia measures and have at best only included SpO2 nadir, an imperfect reflection of nocturnal hypoxia. Ultimately it may be the degree of hypoxia in OSA, rather than AHI, which has the greatest influence on VO2 max. It seems likely that if untreated OSA affects VO2 max it will be most apparent in those with significant oxygen desaturation (expressed by time with SpO2 nadir < 80%).As noted previously, significant variability in demographics including age and sex as well as limited data on degree of oxygen desaturation make it difficult to establish trends between those studies which have shown a reduction in VO2 max and those which have not. Despite this, we note that prior studies showing a reduction in VO2 max in patients with OSA have predominantly involved middle-aged male patients with moderate to severe OSA. In these studies, degree of hypoxia is expressed predominantly by SpO2 nadir and VO2 max reduction has ranged from 3 to 9 mL⋅kg−1⋅min−1 (Table 1). In one of the earliest studies, Vanuxem et al found VO2 max to be reduced by a mean of 6.8 mL⋅kg−1⋅min−1 in middle-aged males with moderate OSA (AHI 25 events/h, SpO2 nadir 85%) compared to controls (26.4 vs 33.2 mL⋅kg−1⋅min−1; P < .005).38 In similarly matched populations, Beitler et al and Lin et al found comparable results in patients with severe OSA. Beitler at al noted a 6.1 mL⋅kg−1⋅min−1 reduction in VO2 max in patients with OSA (AHI 37 events/h, SpO2 nadir 81%), while Lin et al found a 8.5 mL⋅kg−1⋅min−1 reduction in patients with severe OSA with more substantial oxygen desaturation (AHI 44 events/h, SpO2 nadir 65%).62,63 Furthering this trend, Nanas et al found a reduction in VO2 max of 6 mL⋅kg−1⋅min−1 (28.7 vs 34.7 mL⋅kg−1⋅min−1; P < .01) in a study of middle-aged males with moderate to severe OSA (AHI 55 events/h, SpO2 nadir NR), though degree of hypoxia was not reported.50 Studies by Chien et al and Vanhecke found slightly lower reductions of 3.6 mL⋅kg−1⋅min−1 and 3.5 mL⋅kg−1⋅min−1, respectively, in patients with severe OSA, though degree of hypoxia was again not mentioned in these studies.57,64Table 1 Studies showing reduced VO2 max in untreated OSA.StudyMethodsPatient CharacteristicsKey Findings and ResultsVanuxem et al, 199738• Case-control• Type 1 PSG• Cycle ergometer CPET to exhaustion with respiratory gas exchange measurements by a metabolic cart• AHI >10 events/h vs AHI ≤ 10 events/h• Setting: community• 20 patients• Age: 47.8 y (OSA) vs 41.9 y (controls); P > .05• BMI: 26.6 (OSA) vs 26.4; P > .01• All male• AHI: 25.6 events/h (OSA)• Minimum SpO2: 85.4% (OSA)• VO2 max, mL⋅kg−1⋅min−1: 26.2 (OSA) vs 33.2; P < .005• Peak HR: 160 bpm (OSA) vs 166 bpm; P > .05• Peak systolic BP: 206 (OSA) vs 194; P > .05• Peak diastolic BP: 104 (OSA) vs 92; P < .05Lin et al, 200663• Case-control• Type 1 PSG• Cycle ergometer CPET to exhaustion with respiratory gas exchange measurements by a metabolic cart• AHI > 30 events/h vs AHI < 10 events/h• Setting: sleep clinic• 40 patients• Age: 47 y (OSA) vs 44 y (controls); P > .05• BMI: 28.3 (OSA) vs 27.6; P > .05• Male: 90% (OSA) vs 90%; P > .05• AHI: 44 events/h (OSA) vs 5 events/h; P < .05• Minimum SpO2: 65.5% (OSA) vs 91.9%; P < .05• VO2 max, mL⋅kg−1⋅min−1: 21.6 (OSA) vs 30.1; P < .05• Peak HR: 156 bpm (OSA) vs 161 bpm; P > .05Beitler et al, 201462• Case-control• Type 1 PSG• Cycle ergometer CPET to exhaustion with respiratory gas exchange measurements by a metabolic cart• AHI ≥ 15 events/h vs AHI < 15 events/h• Setting: hospital sleep clinic• 34 patients• Age: 47.9 y (OSA) vs 34.3 y (controls); P < .01• BMI: 32.2 (OSA) vs 28.8; P = .17• Male: 80% (OSA) vs 53%; P = .15• AHI: 37.6 events/h (OSA) vs 1.5 events/h; P < .01• Minimum SpO2: 81% (OSA) vs 92%; P < .01• TST SpO2 < 90%: 21.9 (OSA) vs 0 min; P < .01• VO2 max (% predicted): 70.1% (OSA) vs 83.8%; P = .02• VO2 max, mL⋅kg−1⋅min−1: 19.1 (OSA) vs 25.2; P = .04• Peak HR (% predicted): 85 (OSA) vs 90; P = .35• Peak systolic BP: 179 (OSA) vs 173; P = .47• Peak diastolic BP: 81 (OSA) vs 76; P = .18Nanas et al, 201050• Case-control• Type 1 PSG• Treadmill CPET to exhaustion with respiratory gas exchange measurements by a metabolic cart• AHI ≥ 25 events/h vs AHI < 5 events/h• Setting: sleep clinic• 31 patients• Age: 48 y (OSA) vs 46 y (controls); P > .05• BMI: 29.3 vs 28.1; P > .05• All male• AHI: 55 events/h (OSA)• Minimum SpO2: not reported• VO2 max (% predicted): 88% (OSA) vs 98%; P < .05• VO2 max, mL⋅kg−1⋅min−1: 28.7 (OSA) vs 34.7; P < .01• Peak HR (% predicted): 96 (OSA) vs 103; P > .05• Peak HR: 155 bpm (OSA) vs 172 bpm; P > .05Chien et al, 201264• Case-control• Type 1 PSG• Cycle ergometer CPET to exhaustion with respiratory gas exchange measurements by a metabolic cart• AHI ≥ 30 events/h vs AHI < 5 events/h• Setting: sleep clinic• 60 patients• Age: 50 y (OSA) vs 50 y (controls); P = .74• BMI: 26.5 vs 25.8; P = .30• All male• AHI: 48.4 events/h (OSA) vs 2.7 events/h; P < .001• Minimum SpO2: 70% (OSA) vs 86%; P < .001• %TST SpO2 < 90%: 20.6 (OSA) vs 0.1; P < .001• VO2 max, mL⋅kg−1⋅min−1: 25.0 (OSA) vs 27.7; P = .003• Peak HR: 153 bpm (OSA) vs 165 bpm; P = .001Vanhecke et al, 200857• Prospective observational• Type 1 PSG• Treadmill CPET to exhaustion with respiratory gas exchange measurements by a metabolic cart• AHI > 15 events/h vs AHI < 5 events/h• Setting: bariatric clinic• 92 patients• Age: 46 y (OSA) vs 45 y (controls); P = .59• BMI: 50 (OSA) vs 47; P = .15• Male: 32% (OSA) vs 30%; P = .83• AHI: 32.5 events/h (OSA) vs 2.5 events/h; P < .001• Minimum SpO2: not reported• VO2 max, mL⋅kg−1⋅min−1: 17.6 (OSA) vs 21.1; P < .001• % rise HR: 79 (OSA) vs 99; P = .02• % rise systolic BP: 44 (OSA) vs 41; P = .77• % rise diastolic BP: 10 (OSA) vs 2; P = .005Evans et al, 201471• Prospective observational• Type 1 PSG• Cycle ergometer CPET to exhaustion with respiratory gas exchange measurements by a metabolic cart• AHI ≥ 1 event/h vs AHI < 1 event/h• Setting: pediatric hospital sleep clinic• 71 patients• Age: 10 y (OSA) vs 10 y (controls); P = .85• Male: 68% (OSA) vs 35%; P = .81• AHI: 8.9 events/h (OSA) vs 0.4 events/h; P < .001• Minimum SpO2: 85% (OSA) vs 90%; P = .002• VO2 max, mL⋅kg−1⋅min−1: 20.8 (OSA) vs 29.6; P < .001• Peak HR: 159 bpm (OSA) vs 173 bpm; P = .006AHI = apnea-hypopnea index, BMI = body mass index, BP = blood pressure, bpm = beats per minute, CPET = cardiopulmonary exercise testing, HR = heart rate, OSA = obstructive sleep apnea, PSG = polysomnography, SpO2 = oxygen saturation, TST = total sleep time, VO2 = maximum oxygen uptake.Table 2 Studies showing no reduction in VO2 max in untreated OSA.StudyMethodsPatient CharacteristicsKey Findings and ResultsRizzi et al, 201365• Case-control• Type 1 PSG• Treadmill CPET to exhaustion with respiratory gas exchange measurements by a metabolic cart• AHI ≥ 10 events/h vs AHI < 5 events/h• Setting: sleep clinic• 115 patients• Age: 53.7 y (lean OSA) vs 50.8 y (lean controls) vs 50.7 y (obese OSA) vs 49.1 y (obese controls); P = .1• BMI: 22.1 (lean OSA) vs 22.8 vs 33.6 (obese OSA) vs 33.4; P < .01• 31% male• AHI: 22.4 events/h (lean OSA) vs 2.8 events/h (lean controls) vs 33.3 events/h (obese OSA) vs 2.9 events/h (obese controls); P < .01• Minimum SpO2: 86.5% (lean OSA) vs 91.2% (lean controls) vs 78.3% (obese OSA) vs 88.9% (obese controls); P < .01• %TST SpO2 < 90%: 10.5% (lean OSA) vs 0% vs 32.9% (obese OSA) vs 0.2%; P < .01• VO2 max, mL⋅kg−1⋅min−1: 32.1 (lean OSA) vs 30.5 vs 21.7 (obese OSA) vs 24.7; P < .01 (for obesity only, not OSA)• Peak HR: 158 bpm (lean OSA) vs 161 bpm vs 151 bpm (obese OSA) vs 159; P = .1• Peak systolic BP: 174 (lean OSA) vs 173 vs 184 (obese OSA) vs 192; P = .07• Peak diastolic BP: 81 (lean OSA) vs 83 vs 92 (obese OSA) vs 89; P = .02 (obese OSA only)Powell et al, 201967• Case-control• Type 1 PSG• Treadmill CPET to exhaustion with respiratory gas exchange measurements by a metabolic cart• AHI ≥ 15 events/h vs AHI < 15 events/h• Setting: military treatment facility (active-duty personnel with dyspnea)• 98 patients• Age: 40.7 y (OSA) vs 39.4 y (controls); P = .45• BMI: 30.4 (OSA) vs 29.9; P = .46• Male: 97.5% (OSA) vs 93.1%; P = .64• AHI: 32.7 events/h (OSA) vs 5.8 events/h; P < .0001• Minimum SpO2: 84.2% (OSA) vs 88.3%; P = .0008• VO2 max, mL⋅kg−1⋅min−1: 34.9 (OSA) vs 35.5; P = .65• VO2 max (% predicted): 101% (OSA) vs 102%; P = .60• Peak HR: 166 bpm (OSA) vs 171 bpm; P = .09• Peak systolic BP: 178 (OSA) vs 180; P = .77• Peak diastolic BP: 78 (OSA) vs 74; P = .28Hargens et al, 200868• Case-control• Type 3 OCST• Cycle ergometer CPET to exhaustion with respiratory gas exchange measurements by a metabolic cart• AHI ≥ 5 events/h vs AHI < 5 events/h• Setting: community• 44 patients• Age: 22.4 y (OSA) vs 21.4 y (no OSA) vs 21.4 y (controls); P > .05• BMI: 32.0 (OSA) vs 31.4 (no OSA) vs 22.0; P < .05• All male• AHI: 22.7 events/h (OSA) vs 2.5 events/h (no OSA) vs 2.0; P < .05• Minimum SpO2: 86.2 (OSA) vs 88.3 (no OSA) vs 90.0; P > .05• VO2 max, mL⋅kg−1⋅min−1: 27.1 (OSA) vs 28.0 (no OSA) vs 33.2; P < .05 (only for normal BMI controls)• Peak HR: 179 bpm (OSA) vs 180 bpm (no OSA) vs 181 bpm; P > .05• Peak systolic BP: 196 (OSA) vs 202 (no OSA) vs 193; P > .05• Peak diastolic BP: 90 (OSA) vs 91 (no OSA) vs 89; P > .05Barbosa et al, 201869• Case-control• Type 1 PSG• Treadmill CPET to exhaustion with respiratory gas exchange measurements by a metabolic cart• AHI ≥ 15 events/h vs AHI < 5 events/h• Setting: community older adult patients• 28 patients• Age: 70.6 y (OSA) vs 69.3 y (controls); P > .05• BMI: 26.2 (OSA) vs 27.0; P > .05• Male: 23% (OSA) vs 13%; P > .05• AHI: 24.4 events/h (OSA) vs 2.3 events/h; P < .01• Minimum SpO2: not reported• ODI: 14.1 events/h (OSA) vs 2.5 events/h; P < .01• VO2 max, mL⋅kg−1⋅min−1: 17.2 (OSA) vs 16.9; P > .05• Peak HR: 144 bpm (OSA) vs 150 bpm; P > .05• Peak systolic BP: 201 (OSA) vs 197; P > .05• Peak diastolic BP: not reportedCepeda et al, 201566• Case-control• Type 1 PSG and cycle ergometer CPET to exhaustion with respiratory gas exchange measurements by a metabolic cart• AHI ≥ 15 events/h vs AHI < 15 events/h• Setting: outpatient cardiac clinic; patients with metabolic syndrome• 76 patients• Age: 49 y (MetS + OSA) vs 46 y (MetS, no OSA) vs 46 y (controls); P > .05• BMI: 32 (MetS + OSA) vs 32 (MetS, no OSA) vs 25; P < .05• Male: 60% (MetS + OSA) vs 46% (MetS, no OSA) vs 44%; P > .05• AHI: 42 events/h (MetS + OSA) vs 7 events/h (MetS, no OSA) vs 4 events/h; P < .05• Minimum SpO2: 77% (MetS + OSA) vs 88% (MetS, no OSA) vs 91%; P < .05• VO2 max, mL⋅kg−1⋅min−1: 22.6 (MetS + OSA) vs 23.6 (MetS, no OSA) vs 28.7; P < .05 for controls; no difference between MetS groups• Peak HR: 156 bpm (MetS + OSA) vs 164 bpm (MetS, no OSA) vs 166 bpm; P > .05AHI = apnea-hypopnea index, BMI = body mass index, BP = blood pressure, bpm = beats per minute, CPET = cardiopulmonary exercise testing, HR = heart rate, MetS = metabolic syndrome, OCST = out of center sleep testing, ODI = oxygen desaturation index, OSA = obstructive sleep apnea, SpO2 = oxygen saturation, PSG = polysomnography, TST = total sleep time, VO2 = maximum oxygen uptake.Comparatively, studies which have not shown a reduction in VO2 max have involved more diverse OSA populations from age and sex composition but have been similarly limited in their evaluation of nocturnal hypoxia (Table 2). Several of these studies have involved younger adults and geriatric populations as well as a significantly larger representation of female patients. In a study by Rizzi et a