Muscle oxygen extraction-utilization limitation (5) has been identified as a mediator of diminished aerobic capacity (19,20,23,25) in individuals who are infected with the human immunodeficiency virus (HIV). Highly active antiretroviral therapy (HAART), the medication regimen used to retard the progression of HIV infection, has been known to disrupt normal mitochondrial function (30), making it a likely contributor to muscle oxygen extraction-utilization limitation in most individuals treated for HIV infection. Nucleoside analogs, the principal agents of HAART, have been reported to be DNA chain terminators and their side effects included attenuated expression of cytochrome c oxidase (4,8), inhibition of NADH-linked respiration and NADH-cytochrome c reductase activity (22), pathologic changes in the mitochondrial structure, and damaged and decreased mitochondrial DNA and RNA (3,10). These alterations in mitochondrial function may have resulted in decreased oxygen utilization, even when delivery of oxygen to the muscle was unaffected (5).
Arteriovenous oxygen difference (a-vO2) has been well accepted as a global indicator of muscle oxygen extraction-utilization when measured during exercise (1). We have previously observed decreased peak oxygen consumption accompanied by diminished peak a-vO2 during maximal incremental treadmill exercise in a group of individuals with HIV taking HAART compared with an age, gender, and physical activity level matched non-HIV infected control group (5). If the untoward effects of HAART resulted in reduced muscle oxygen extraction-utilization, then, during exercise, a-vO2 would be lower in individuals with HIV taking HAART than in those with HIV but not taking HAART. If the untoward effects of the HIV virus resulted in reduced muscle oxygen extraction-utilization during exercise, a-vO2 would be lower in both individuals with HIV taking and not taking HAART. The aim of this study was to compare central oxygen delivery (cardiac output) and peripheral muscle oxygen extraction-utilization (a-vO2) in individuals with HIV taking and not taking HAART during incremental maximal treadmill exercise.
Subjects participating in this study were 15 individuals infected with HIV taking HAART, 15 individuals with HIV not taking HAART, and 15 non-HIV infected healthy controls. Seventy-five subjects were screened for inclusion in the study. All subjects were matched for age, gender, and activity level (Table 1) and were considered sedentary, defined as not having participated in a routine exercise program in the past 6 months or employed in an occupation that induces perspiration on the average of once per week, discounting the effects of ambiance. Subjects infected with HIV were not diagnosed with acquired immune deficiency syndrome (AIDS): having CD4 counts equal to or greater than 200 cells·mm−3 (Table 2), no opportunistic infections, and were considered asymptomatic (Centers for Disease Control and Prevention Classification A1 or A2) (6). Subjects infected with HIV taking HAART were on their medication regimen for a minimum of six months with self-reporting compliance to be high. The majority of subjects with HIV not taking HAART (12 of 15) were totally naïve to HAART, and no subject had taken HAART for a minimum of 6 months before beginning the study (Table 3). Nucleoside analog induced mitochondrial myopathy has been shown to reverse after cessation of the drug for three months (21). No subject in any of the groups was taking medication, with the exception of HAART, that would affect the cardiorespiratory response to exercise. No subject had a known cardiorespiratory, neurological, pulmonary, hematological, renal or orthopedic condition, other than HIV, that would limit or affect the ability to perform treadmill exercise. Written consent from each subject was obtained before exercise testing according to the guidelines issued by the university’s institutional review board.
All maximal exercise tests were performed on a Trackmaster® treadmill. Metabolic data was continuously measured by a MedGraphics® metabolic cart, which was calibrated using gas mixtures of 21% oxygen and 0% carbon dioxide and 12% oxygen and 5% carbon dioxide, both in nitrogen balances. Zirconium cell oxygen and an infrared cell carbon dioxide analyzers were used to measure the percentages of inspired and expired oxygen and carbon dioxide. Inspired and expired ventilatory volume data were obtained via flow volume loops generated by a flow-sensitive pneumotachometer interfaced with a microprocessor. Calibration of the pneumotachometer was performed using a 3-L syringe with injections at various flow rates. Noninvasive cardiac output (𝑄̇t) was measured using the exponential rise carbon dioxide rebreathing technique originally described by Defares (11). The rebreathing apparatus consisted of a 98.0 mL Hans Rudolph® slide valve mounted on a headset, an attached mouthpiece, and a 5.0-L anesthetic bag used for rebreathing, and a vacuum sample line attached to the metabolic cart. A volume of 1.5–2.0 times the subject’s tidal volume (Vt) (17) of concentrated gas (4% carbon dioxide, 35% oxygen, balanced nitrogen) was injected into the rebreathing bag. A three-way stopcock was attached to the rebreathing bag and to a 3.0-L syringe for the purpose of controlling the injected gas mixture into the rebreathing bag. This syringe was further connected to a tank, containing the rebreathing gas mixture. Heart rate (HR) and rhythm were measured using 12-lead electrocardiography (ECG). Cardiac stroke volume (Vs) was determined as the ratio of 𝑄̇t and HR. Functional aerobic impairment was defined as a peak oxygen consumption (V̇O2) ≥27% below predicted values based on age, gender, and physical activity (13). The Aerobic Impairment Index (AII), used to determine the presence of functional aerobic impairment, was determined by the equation: AII = (expected oxygen consumption − measured oxygen consumption)/(expected oxygen consumption) × 100. Expected peak V̇O2 in milliliters per kilogram per minute was given by: 42.3–0.356 × (age in years) for sedentary females and 57.8–0.445 × (age in years) for sedentary males (13). Arterialized blood lactic acid concentration was measured by fingerstick lactate analysis (Accusport Inc.®). Subjects who smoked were instructed to refrain from smoking for at least 5 h before the exercise test to control for the effects the effects of acute carboxyhemoglobinemia. To evaluate ventilatory economy, ventilatory quotients (V̇E/V̇O2) for peak V̇O2 and (V̇E/V̇CO2) for peak V̇CO2 were calculated for all subjects. Body surface area (BSA) was determined by the equation: 0.20247 × height (m)0.725 × weight (kg)0.425, and body mass index (BMI) was determined by the equation: weight (kg)/height(m2).
Sample size was calculated from preliminary work performed in our laboratory (5). In this preliminary study, 15 subjects with HIV taking HAART for at least 6 months, and 15 age, gender, and activity level matched non-HIV infected controls performed a maximal incremental treadmill exercise test to exhaustion. Cardiac output and a-vO2 were measured at each stage and peak exercise as described in the present study. Means and standard deviations of peak a-vO2 for participants with HIV and controls were 10.8 ± 1.9 vol% and 12.4 ± 1.9 vol%, respectively. These values were significantly different at the P < 0.05 level. There were no significant differences in peak cardiac output. The sample size was calculated as 11.8 subjects per group. In the present study, all subjects completed a maximal treadmill exercise test using the modified Bruce protocol (2). For the exercise test to be considered maximal, subjects were required to reach at least one of the following criteria: attainment of ≥ 90% of their age-predicted maximal HR (220 − age), a peak respiratory exchange ratio of ≥ 1.10, or a peak lactate ≥ 8 mmol (2). Subjects not reaching at least one required criterion were excluded from the study. Before beginning the test, silver/silver chloride ECG electrodes were placed on the subject’s chest in the Mason-Likar configuration (24). The subjects then donned the rebreathing apparatus and a nose-clip was positioned to ensure that no air would escape through the nose. After resting finger-stick lactate analysis, the subjects transferred to the standing position on the treadmill in order to perform rebreathing measurements during standing rest. With the slide valve in the open position, subjects inspired room air for measurement of resting oxygen consumption (V̇O2), carbon dioxide expiration (V̇CO2), and partial pressure of end tidal carbon dioxide (PETCO2). The rebreathing bag was then filled with the 4% carbon dioxide and 35% oxygen gas mixture, and the slide valve was closed to perform the rebreathing procedure. During the rebreathing procedure, the subjects repeatedly inspired and expired the gas mixture contained in the rebreathing bag. Rebreathing did not continue for more than 6 s to prevent overestimation of PCO2 due to recirculation (9,15). Subjects were asked to breathe at the rate of 40 breaths per minute, as dictated by an auditory metronome, to ensure acquisition of an adequate amount of data points for use in plotting the exponential rise in expired carbon dioxide. After acquisition of the rebreathing data, the slide valve was returned to the open position, and the subjects again breathed room air. Mixed venous PCO2 was subsequently estimated from the asymptote of the exponential rise in the end tidal PCO2 obtained during the rebreathing procedure (11). An iterative technique was then applied to log (PV̇CO2 − PETCO2) in order to reduce the variance of the values of PV̇CO2 around the least squares regression line as described by Heigenhauser and Faulkner (14). Arterial carbon dioxide was obtained using an algorithm described by Jones et al. (18): PaCO2 = 5.5 + 0.90 PETCO2 −0.0021 tidal volume (Vt). Venous-arterial content difference in CO2 (Cv-aCO2) was then calculated using standard carbon dioxide dissociation curves expressed by the equations: Log CV̇CO2 = (logPV̇CO2 × 0.396) + 2.38 and Log CaCO2 = (logPaCO2 × 0.396) + 2.38. Cardiac output was then determined from the estimated Cv-aCO2 and measured V̇CO2 by the equation: pulmonary circulation (𝑄̇l) = 𝑄̇t = V̇CO2/Cv-aCO2 (16). After acquisition of the resting data in the standing position on the treadmill, subjects began the exercise test. Speed and elevation was advanced according to the modified Bruce protocol (2). The CO2 rebreathing procedure was repeated at the end of the second minute of each 3-min stage and at peak exercise. A second finger stick lactate analysis was performed at 1.5 min after volitional exhaustion. a-vO2 was calculated as the ratio V̇O2/𝑄̇t and reported in vol% units. Vs was calculated as the ratio 𝑄̇t/HR and in reported in mL.
Pearson product moment correlation coefficients were calculated to identify relationships among dependent and demographic variables. ANOVA were used to determine whether demographic variables including age, height, weight, BMI, BSA, CD4 count, viral load, hematocrit, hemoglobin, white blood cell count, and time since HIV diagnosis were significantly different between the groups. Two-way ANOVA, with repeated power output in the first order and intergroup comparison in the second order, were used to determine whether there were significant differences in V̇O2, 𝑄̇t, a-vO2, Vs, and HR as a result of increases in treadmill power output. Interaction of the first- and second-order effects was analyzed to determine whether intergroup differences occurred at individual power outputs. Separate ANOVA for each group were used to identify peak V̇O2 plateaus by determining whether V̇O2 significantly differed between the last two work stages. One-way ANOVA were used to identify significant intergroup differences in peak cardiorespiratory variables. Both statistically and biologically plausible covariance resulted from relationships among peak V̇O2, BSA, and 𝑄̇t. Subsequently, data were analyzed using ANCOVA with data adjusted for the interaction among peak V̇O2, BSA, and 𝑄̇t. To determine the effects of gender and race on the dependent variables, three-way ANOVA were computed. Main effects included gender, race, and group (HIV+ on HAART, HIV+ not on HAART, and control). Interactions included gender by race, gender by group, race by group, and gender by race by group.
Significant relationships and differences in the means and the means associated with the secondary analyses were identified when the probability of making a Type-I error was equal to or less than five percent (P ≤ 0.05). Data were reported as means plus or minus one standard error (SE) throughout the tables and text.
Groups differed in demographic characteristics only in the ratio of African-Americans to Caucasians in each group (Table 1). Serological characteristics in groups with HIV were similar with the exception of a greater viral load in those subjects with HIV not taking HAART (Table 2). The medication regimen used by subjects with HIV taking HAART is presented in Table 3. Route of HIV infection for groups with HIV are presented in Table 4. There were no significant differences among the cardiorespiratory variables at rest except 𝑄̇t was significantly lower (P < 0.05) in those with HIV taking HAART than in controls (Table 5).
General exercise responses.
All subjects infected with HIV and controls stopped exercising due to volitional exhaustion despite strong encouragement to continue and no subject reported another reason for stopping exercise. All subjects reached at least stage 5 of the modified Bruce treadmill protocol, while many reached higher stages (Table 6). Because all subjects reached at least stage 5 but some did not exceed stage 5 of the exercise protocol, data in Figure 1 are presented as rest, stages 1–4, and the peak stage of all groups. V̇O2, 𝑄̇t, and HR increased as expected with increases in power output (Figs. 1 and 2) in all three groups (P < 0.05). Vs and a-vO2 also rose during exercise (Figs. 1 and 2). Peak oxygen consumption was not significantly different between the last two stages of the exercise tests and all subjects reached an RER of at least 1.15 (Table 7). Peak lactate concentration exceeded 8.0 mmol·dL−1 only in the control group and was significantly lower (P < 0.05) in both groups with HIV than in the controls (Table 7).
Peak aerobic capacity.
Viral load, CD4 count, white blood cell count, hematocrit, hemoglobin levels (Table 2), the time (months) since HIV infection, and the duration (months) of HAART (Table 3) were not significantly related to peak V̇O2, HR, 𝑄̇t, Vs, or a-vO2. With all groups combined and also when subjects with HIV and controls were examined independently, V̇O2 during all exercise stages was significantly (P < 0.05) related to 𝑄̇t, Vs, HR, and a-V̇O2. Peak oxygen consumption was significantly lower (P < 0.05) in subjects with HIV taking HAART than in controls (Table 7). There was a trend (P < 0.06) toward lower peak V̇O2 in the group with HIV not taking HAART compared with controls but was not significant. There was no difference in peak V̇O2 between groups with HIV. Ten of 15 subjects with HIV taking HAART (AII = 29.9 ± 2.8%), 8 of 15 subjects with HIV (AII = 24.1 ± 3.8%) not taking HAART, and 1 of 15 controls (AII = 3.2 ± 4.6%) had functional aerobic impairment. 𝑄̇t and Vs at peak exercise were not significantly different among those with HIV taking HAART, those with HIV not taking HAART, and controls (Table 7). Peak HR was significantly lower (P < 0.05) in subjects with HIV taking HAART than in those with HIV not taking HAART and controls (Table 7). Compared with subjects with HIV not taking HAART and controls, the a-vO2 response was blunted in the subjects with HIV taking HAART (Fig. 2). Peak a-vO2 difference was significantly lower (P < 0.05) in subjects with HIV taking HAART than in subjects with HIV not taking HAART and controls (Table 7). Peak oxygen pulse was significantly lower (P < 0.05) in those with HIV taking HAART than controls. There was no difference in ventilatory quotients between groups.
Peak aerobic capacity adjusted for covariance.
Due to the possibility of a dissimilar peak V̇O2 values in the groups with HIV taking HAART compared with the group with HIV not taking HAAAT and the control group accounting for the lower a-vO2 in the group with HIV taking HAART, we decided to adjust for covariance in peak V̇O2 and other peak metabolic parameters in a secondary analysis. Peak V̇O2 expressed both as milliliters per minute and milliliters per kilogram per minute was lower in subjects with HIV taking HAART compared with controls (P < 0.0001) when the analyses were adjusted for covariance due to the interaction of BSA and 𝑄̇t (Table 8). Peak cardiac output remained similar among the groups when adjusted for covariance with BSA and peak V̇O2. Peak a-vO2 remained significantly lower in subjects with HIV taking HAART compared with subjects with HIV not taking HAART (P < 0.02) and controls (P < 0.0003) when the analysis was adjusted for covariance with the interaction of BSA, peak V̇O2, and peak 𝑄̇t (Table 8). Peak Vs was significantly higher in subjects with HIV taking HAART compared with subjects with HIV not taking HAART (P < 0.005) and controls (P < 0.0003) when adjusted for covariance with the interaction of BSA, peak V̇O2, and peak 𝑄̇t (Table 8). The adjusted peak HR was significantly lower in subjects with HIV taking HAART than in the group with HIV not taking HAART (P < 0.05) and controls (P < 0.02) adjusted for covariance with the interaction of BSA, peak V̇O2, and peak 𝑄̇t (Table 8). Peak exercise stage was also significantly lower in subjects with HIV taking HAART than in the other groups (P < 0.03) when adjusted for covariance with peak V̇O2, BSA, and 𝑄̇t.
Peak V̇O2 and a-vO2 were also examined adjusting for covariance in duration of HIV infection, despite no association with any metabolic variable or significant differences between groups with HIV, because of the large standard error between groups with HIV. Peak V̇O2 remained significantly lower (P < 0.05) in those with HIV taking HAART compared with controls and peak a-vO2 remained significantly lower (P < 0.025) in those taking HAART compared with those with HIV not taking HAART and controls.
Effects of gender and race.
As expected, there were significant gender differences (P < 0.0001) in peak V̇O2 (33.4 ± 2.1 vs 25.3 ± 1.7 mL·kg−1·min−1), 𝑄̇t (22.2 ± 1.0 vs 15.7 ± 0.9 L·min−1), and Vs (124.4 ± 5.6 vs 88.6 ± 5.2), with males exhibiting the greater values for all variables. There were also significant differences between races for relative peak V̇O2 (African-American = 27.2 ± 1.9 mL·kg−1·min−1 vs Caucasian = 36.2 ± 1.8 mL·kg−1·min−1; P < 0.03) and a-vO2 (African-American = 10.6 ± 0.4 vol% vs Caucasian = 13.2 ± 0.5 vol%; P < 0.01). Significant differences for interactions of gender, race, and group were not observed. When analyzed independently (Table 9), there was no significant difference in peak V̇O2 for African-Americans with HIV on HAART and not on HAART. Peak a-vO2 remained significantly lower (P < 0.025) in African-Americans with HIV taking HAART than in African-Americans with HIV who were not taking HAART. There were no significant differences in peak 𝑄̇t, HR, or Vs between African-Americans taking and not taking HAART (Table 9).
Description of mechanistic differences in peak aerobic capacity.
Significant differences in peak a-vO2 (Tables 7 and 8) were observed between HIV-infected subjects taking HAART and HIV-infected subjects not taking HAART and controls. A significant difference in a-vO2 between those with HIV who were not taking HAART and controls was not observed. These findings were accompanied by a lower peak HR in those with HIV taking HAART than in the other remaining groups. Because peak HR was significantly lower in the group with HIV taking HAART compared with those with HIV not taking HAART and controls, we found biological plausibility to examine exercise variables adjusting for covariance in peak HR. When adjusted for covariance with peak HR, peak a-vO2 remained significantly lower in subjects with HIV taking HAART (9.9 ± 0.5 vol%) than subjects with HIV not taking HAART (11.6 ± 0.5 vol%) (P < 0.02) and controls (12.9 ± 0.5 vol%) (P < 0.0002). Differences in peak HR were no longer present when covariance with a-vO2 was normalized. When adjusting for covariance due to the interaction of peak V̇O2, peak HR, and BSA, peak Vs was significantly higher (P < 0.05) in those with HIV taking HAART than in those with HIV not taking HAART and controls.
Results of this study were in agreement with previous reports that have underscored diminished aerobic capacity in individuals infected with HIV (5,19,20,23,25). In a previous study, decreased muscle oxygen extraction-utilization was determined to be a mediator of diminished aerobic capacity in individuals with HIV infection, but this work included only subjects infected with HIV who were taking HAART (5). Results of the current study have expanded the available information by determining that decreased muscle oxygen extraction-utilization was present only in individuals with HIV who are taking HAART and not in individuals with HIV who are not taking HAART. Findings of the current study have implicated HAART as a primary contributor to decreased muscle oxygen extraction-utilization in individuals infected with HIV who are routinely treated for this infection.
The components of the HAART regimen usually have included at least one nucleoside analog, a nonnucleoside analog reverse transcriptase inhibitor, and/or a protease inhibitor. An association between protease inhibitors and impaired glucose and lipid metabolism has been previously reported (29). A tertiary analysis of the current data revealed that a-vO2, and therefore peripheral muscle oxygen extraction-utilization, was significantly lower (P < 0.05) in subjects on a HAART regimen that included nucleoside analogs but did not include a protease inhibitor (N = 10, 6 males, 4 females; 9.5 ± 0.9 vol%) compared with those who were using a protease inhibitor (N = 5, 3 males, 2 females; 10.7 ± 1.2 vol%), suggesting no effect of protease inhibitors on muscle oxygen extraction-utilization. Deleterious effects on metabolism have not been reported in the literature as a result of nonnucleoside reverse transcriptase inhibitor use. In the current study, there was no difference in a-vO2 in groups taking (N = 7) and not taking (N = 8) nonnucleoside reverse transcriptase inhibitors. Use of nucleoside analogs has been known to inhibit HIV replication by substitution of the phosphorylated drug into the DNA in place of deoxy-thymidine triphosphate (dTTP), resulting in termination of the elongating DNA chain in the 3′ to 5′ direction, thereby blocking reverse transcription. Nucleoside analogs have a reported similar effect on the mitochondrial transcription enzymes, DNA polymerase γ, and mitochondrial DNA chain terminators (26). These alterations in mitochondrial function (22) are thought to impair oxidative metabolism (30) and result in decreased muscle oxygen extraction-utilization. Consequently, nucleoside analogs appeared to be the most probable agents of diminished muscle oxygen extraction-utilization in the HAART regimen.
Attainment of maximal aerobic capacity during treadmill testing has been identified as the attainment of a respiratory exchange ratio (RER) of 1.10 or greater, a peak heart rate of at least 90% of the age-predicted maximum heart rate, a peak lactic acid concentration of at least 8.0 mmol·dL−1, and/or observation of an oxygen consumption plateau resulting from a lack of increase in V̇O2, despite a substantial increase in power output. In the current study, a significant rise in V̇O2 was not observed with increased power output during the last two work stages of the treadmill test in any of the groups. Despite a significantly lower peak heart rate in the group of subjects with HIV taking HAART, mean peak HR exceeded 90% of the age-predicted maximum heart rate in all of the groups. Ten of 15 subjects in the group with HIV taking HAART, 11 of 15 subjects in the group with HIV not taking HAART, and all of the controls reached at least 90% of the age-predicted maximum HR. All subjects in the current study attained a RER of at least 1.15. Upon performing a secondary analysis adjusting for covariance in peak HR, therefore, normalizing HR among groups, peak V̇O2 and peak a-vO2 remained significantly lower in those with HIV taking HAART compared with controls. These findings indicated that maximal aerobic capacity was attained at volitional exhaustion and as a test end point. Malingering or other nonphysiological reasons were not probable confounders to the interpretation of the results. The lower HR at peak exercise in subjects with HIV taking HAART when adjusting for covariance in peak metabolic parameters may have indicated local muscle fatigue before reaching a maximal HR due to a muscle oxygen extraction-utilization limitation (12). The higher Vs at peak exercise in the subjects with HIV taking HAART than those with HIV not taking HAART and controls after adjusting for covariance in peak metabolic parameters may have suggested a central circulatory adaptation to the peripheral extraction-utilization limitation. On the contrary, it is possible to have been the result of a longer diastolic filling time secondary to a lower peak HR in these individuals. However, because peak Vs remained significantly higher after analysis with adjustment for covariance in peak V̇O2, peak HR, and BSA, we felt that this was not probable. With groups collapsed, peak V̇O2, 𝑄̇t, and Vs were higher in males than in females. These differences were in agreement with previous reports (27). Males and females were equally distributed among groups that eliminated confounding of the interpretation due to gender differences. The grouping of subjects according to race was a general category, which included possible bias from associated socioeconomic, educational and other cultural and cross-cultural influences. With the analyses not adjusted for sociodemographic influences, the finding of a higher peak V̇O2 in Caucasians than in African-Americans was in agreement with previous reports (28), but the racial difference did not persist as a function of the interaction of race with gender and group. Caucasians exhibited a higher a-V̇O2 than African-Americans as a main effect, and this difference persisted with subsequent analyses. When only African-Americans with HIV were included in the analysis (Table 9), those taking HAART had significantly lower a-vO2 than those not taking HAART. These findings indicated that HAART exerted an influence on a-vO2 that was independent of influences exerted by gender or racial grouping differences.
Although there were no significant differences between groups in CD4 count, a higher viral load was observed in subjects with HIV not taking HAART than in those on the medication regimen. It was possible that subjects with HIV taking HAART had lower CD4 counts and higher viral loads at some point before they began taking the HAART regimen, which may have indicated a greater disease progression than those not taking HAART. If low immunological function and/or high viral loads caused irreversible structural muscle damage, it was possible that the low a-vO2 seen in subjects with HIV taking HAART was not due to the influence of HAART but rather greater muscle damage that occurred before initiation of the HAART regimen. In the current study, there was a wide range in both CD4 and viral load in both subjects with HIV taking and not taking HAART. Moreover, there was no significant relationship among CD4, viral load, and any of the cardiorespiratory variables, particularly peak a-vO2. The higher viral load observed in those not taking HAART than those taking HAART was not accompanied by a lower a-vO2. Peak a-vO2 remained significantly lower in the group with HIV taking HAART when adjusted for covariance in the duration of HIV infection. The current study excluded subjects diagnosed with clinical AIDS, and subjects with a CD4 count history of less than 200 at any point in the clinical course of the infection (immunological AIDS). As a result, it did not appear that disease progression significantly affected a-vO2 in these subjects with mild asymptomatic HIV infection. However, this must be made with a degree of caution because, although not significantly different, there was also a trend (P < 0.06) toward a lower peak V̇O2 in those with HIV not taking HAART compared with noninfected controls, indicating a potential influence of the HIV virus itself on peak V̇O2. This influence may not have been large enough to produce a significant difference in the present study but potentially may been seen in subjects with HIV who are further along in the disease progression (i.e., symptomatic HIV or AIDS). Moreover, how HIV effects metabolic function during exercise may present differently than the effects of HAART. Further research examining the components of aerobic capacity in individuals at different stages of HIV may reveal this information.
A potential confounder in this analysis is the potential of deleterious side effects that could have caused a more sedentary lifestyle, thus affecting peak V̇O2 and a-vO2. Another potential confounder may be one of the chronic effects of carboxyhemoglobinemia. Long-term effects of smoking may have had an effect of peripheral blood flow during exercise, therefore limiting oxygen extraction in those who smoked. In the current study, approximately 30% of subjects in both groups with HIV smoked regularly. Smoking duration (i.e., pack year) was not determined in these individuals but thought to be similar in both groups with HIV taking and not taking HAART. Because groups with HIV were evaluated independently and peak a-vO2 remained significantly lower in those taking HAART, it seemed the effects of smoking did not exert a significant influence on peak a-vO2.
In the current study, peak lactate production was significantly lower in both groups with HIV compared with controls (Table 7). We are currently unable to interpret this finding. Hypothetically, a limitation in the ability of the muscle to extract and utilize oxygen would have shifted metabolism to a state of anaerobiosis and a state of lactic acidosis would have sooner resulted, possibly to a greater extent, than in an unimpaired muscle. HAART, specifically nucleoside analogs, has been historically associated with resting lactic acidosis. To our knowledge, exercise lactate metabolism in individuals with HIV taking or not taking HAART is unknown. In the present study, both groups with HIV demonstrated lower peak lactate production, suggesting that HIV, rather than HAART, may have affected lactate metabolism during exercise in these individuals. Further work in this area is warranted.
In conclusion, peak a-vO2 was diminished in subjects infected with HIV taking HAART compared with subjects infected with HIV who were not taking HAART and noninfected controls. Subjects with HIV who were not taking HAART had a-vO2 that was not significantly different from controls. Nonnucleoside reverse transcriptase and protease inhibition appeared to have little effect on mitochondrial function. Nucleoside analog medication has been demonstrated to decrease mitochondrial enzyme function (4,7,22). Therefore, the limited ability of the muscles to extract and utilize oxygen may have been the result of diminished oxidative enzyme function as opposed to alterations in other factors that determine a-vO2 such as oxygen convection-diffusion coupling. Further investigations examining muscle oxygen extraction and mitochondrial oxidative enzyme content in individuals infected with HIV taking and not taking HAART were considered warranted.
1. Ades, P. A., M. L. Waldmann, W. L. Meyer, et al. Skeletal muscle and cardiovascular adaptations to exercise conditioning in older coronary patients. Circulation 94: 323–330, 1996.
2. American College of Sports Medicine ACSM’s Guidelines for Exercise Testing, and Prescription, 6th Ed. Baltimore: Lippincott, Williams & Wilkins, 2000, pp. 76–77.
3. Arnaudo, E., M. Dalakas, S. Shanske, C. T. Moraes, S. Dimauro, and E. A. Schon. Depletion of muscle mitochondrial DNA in AIDS patients with zidovudine- induced myopathy. Lancet 337: 508–510, 1991.
4. Benbrik, E., P. Chariot, S. Bonavaud, et al. Cellular and mitochondrial toxicity of zidovudine (AZT), didanosine (ddI) and zalcitabine (ddC) on cultured human muscle cells. J. Neurol. Sci. 149: 19–25, 1997.
5. Cade, W. T., L. E. Fantry, S. R. Nabar, and R. E. Keyser. Decreased peak arteriovenous oxygen difference in HIV-infected individuals. Arch. Phys. Med. Rehabil.
6. Centers for Disease Control and Prevention. Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR 41: RR-17, 1992.
7. Chariot, P., E. Benbrik, A. Schaeffer, and R. Gherardi. Tubular aggregates and partial cytochrome c oxidase deficiency in skeletal muscle of patients with AIDS treated with zidovudine. Acta Neuropathol. 85: 431–436, 1993.
8. Chariot, P., F. Le Maguet, F. J. Authier, D. Labes, F. Poron, and R. Gherardi. Cytochrome c oxidase deficiency in zidovudine myopathy affects perifascicular muscle fibres and arterial smooth muscle cells. Neuropathol. Appl. Neurobiol. 21: 540–547, 1995.
9. Collier, C. Determination of mixed venous CO2
tensions by rebreathing. J. Appl. Physiol. 9: 25–29, 1955.
10. de la Asuncion, J. G., M. L. del Olmo, J. Sastre, et al. AZT treatment induces molecular and ultrastructural oxidative damage to muscle mitochondria: prevention by antioxidant vitamins. J. Clin. Invest. 102: 4–9, 1998.
11. Defares, J. Determination of PvCO2
from the exponential CO2
rise during rebreathing. J. Appl. Physiol. 13: 159–164, 1958.
12. Emmett, J. D. A review of heart rate and blood pressure responses in the cold in healthy subjects and coronary artery disease patients. J. Cardiopulm. Rehabil. 15: 19–24, 1995.
13. Franklin, G. S., S. Gordon, and G. C. Timmis. Fundamentals of exercise physiology: implications for exercise testing and prescription. In: Exercise in Modern Medicine, G. S. Franklin, S. Gordon, and G. C. Timmis (Ed.) Baltimore: Williams & Wilkins, 1989, p. 8.
14. Heigenhauser, G. J., and J. A. Faulkner. Estimation of cardiac output
by the CO2
rebreathing method during tethered swimming. J. Appl. Physiol. 44: 821–824, 1978.
15. Jacob, S. V., L. Hornby, and L. C. Lands. Estimation of mixed venous PCO2
for determination of cardiac output
in children. Chest 111: 474–480, 1997.
16. Jones, N. L. Clinical Exercise Testing, 4th Ed. Philadelphia: W. B. Saunders, 1997, pp. 150–163.
17. Jones, N. L., E. J. Campbell, G. J. Mchardy, B. E. Higgs, and M. Clode. The estimation of carbon dioxide pressure of mixed venous blood during exercise. Clin. Sci. 32: 311–327, 1967.
18. Jones, N. L., D. G. Robertson, and J. W. Kane. Difference between end-tidal and arterial PCO2
in exercise. J. Appl. Physiol. 47: 954–960, 1979.
19. Keyser, R. E., L. Peralta, W. T. Cade, S. Miller, and J. Anixt. Functional aerobic impairment in adolescents seropositive for HIV: a quasiexperimental analysis. Arch. Phys. Med. Rehabil. 81: 1479–1484, 2000.
20. Macarthur, R. D., S. D. Levine, and T. J. Birk. Supervised exercise training improves cardiopulmonary fitness in HIV-infected persons. Med. Sci. Sports Exerc. 25: 684–688, 1993.
21. Masanes, F., A. Barrientos, M. Cebrian, et al. Clinical, histological and molecular reversibility of zidovudine myopathy. J. Neurol. Sci. 159: 226–228, 1998.
22. Modica-Napolitano, J. S. AZT causes tissue-specific inhibition of mitochondrial bioenergetic function. Biochem. Biophys. Res. Commun. 194: 170–177, 1993.
23. Perna, F. M., A. Laperriere, N. Klimas, et al. Cardiopulmonary and CD4 cell changes in response to exercise training in early symptomatic HIV infection. Med. Sci. Sports Exerc. 31: 973–979, 1999.
24. Pina, C. R. Lead systems: sensitivity and specificity. Cardiol. Clin. 2: 329–335, 1984.
25. Pothoff, G., K. Wassermann, and H. Ostmann. Impairment of exercise capacity in various groups of HIV-infected patients. Respiration 61: 80–85, 1994.
26. Simpson, M. V., C. D. Chin, S. A. Keilbaugh, T. S. Lin, and W. H. Prusoff. Studies on the inhibition of mitochondrial DNA replication by 3′-azido- 3′-deoxythymidine and other dideoxynucleoside analogs which inhibit HIV-1 replication. Biochem. Pharmacol. 38: 1033–1036, 1989.
27. Sparling, P. B. A meta-analysis of studies comparing maximal oxygen uptake in men and women. Res. Q. Exerc. Sport 51: 542–552, 1980.
28. Suminski, R. R., R. J. Robertson, F. L. Goss, and S. Arslanian. Peak oxygen consumption and skeletal muscle bioenergetics in African-American and Caucasian men. Med. Sci. Sports Exerc. 32: 2059–2066, 2000.
29. Walli, R., O. Herfort, G. M. Michl, et al. Treatment with protease inhibitors associated with peripheral insulin resistance and impaired oral glucose tolerance in HIV-1-infected patients. AIDS 12: F167–F173, 1998.
30. Weissman, J. D., I. Constantinitis, P. Hudgins, and D. C. Wallace. 31P magnetic resonance spectroscopy suggests impaired mitochondrial function in AZT-treated HIV-infected patients. Neurology 42: 619–623, 1992.