INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Morbidly obese patients are known to have an increased basal metabolic rate (BMR) and rate of total body oxygen consumption (O₂) (1, 2). The percentage of total body oxygen utilization varies among different tissue beds, with adipose tissue demonstrating a lower O₂ than lean tissue such as skeletal muscle and visceral organs. Because adipose tissue has a lower metabolic rate than lean tissue, morbidly obese patients typically have lower total body O₂ levels than nonobese patients when such measures are standardized to body weight (2).

The percentage of cardiac output and total body O₂ dedicated to respiratory muscle work (O_2RESP) during quiet breathing is very small (less than 3%) in nonobese animals and humans (3, 4). O_2RESP has been shown to increase exponentially with increases in work of breathing (5). As minute ventilation (E) increases from normal resting values, the rate of increase in O₂ is significantly greater in obese compared with nonobese patients (6). Changes in posture from sitting to supine are also associated with significant increases in O₂ in obese patients (7). An unproven assumption is that the weight of the abdominal contents and chest wall increase ventilatory load thereby increasing the energy cost of breathing (8). It is speculated that an increased work of breathing related to obesity may predispose these individuals to respiratory failure when acute insults such as airflow obstruction or pneumonia occur.

Despite such speculation, the percentage of total body oxygen consumption dedicated to normal breathing at rest in morbidly obese patients is not known. Accordingly, we measured O₂ in spontaneously breathing, morbidly obese patients immediately prior to scheduled gastric bypass surgery, and again during mechanical ventilation to determine the effects of morbid obesity on O_2RESP during normal spontaneous breathing and compared these values with nonobese control patients.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study was approved by the institutional review board at the University of Chicago and all patients gave informed consent before entering the study. We studied morbidly obese patients scheduled for elective gastric bypass surgery. Morbid obesity was defined as a body mass index (BMI)  40 kg/m². Nonobese patients scheduled for elective abdominal surgery served as control subjects. Baseline demographic data (age, sex, weight, height, BMI, body surface area [BSA]), as well as past medical history and smoking history were obtained for all patients. All patients had baseline fasting O₂ measurements performed in the preoperative holding area using a metabolic monitor (SensorMedics Deltatrac Metabolic Monitor; Datex Instrumentation Corp., Helsinki, Finland). The monitor was calibrated to atmospheric pressure and a standard gas blend of 96% O₂ and 4% CO₂ prior to each use. After a 10-min equilibration period, O₂ was determined as the average of 10 consecutive 1-min measurements. Measurements were made with patients reclined at 30° to 45° from horizontal. All patients were sedated in the preoperative holding area by the anesthesiologist managing the case who was instructed to sedate the patient until he or she was "cooperative, oriented, calm and tranquil" (Ramsay Sedation Scale score of 2) (9). Care was taken to assure that there were no obstructive apneas during O₂ measurements. The choice of preoperative sedative agents was directed by the anesthesiologist who was not involved in the study. At no time during the study was the anesthesiologist in charge of the case aware of the results of the O₂ measurements. O₂ measurements were not begun until each patient had reached the level of preoperative sedation described previously.

In the operating room, general anesthesia was induced and all patients were fully paralyzed with a nondepolarizing neuromuscular blocking agent. The use of anesthetic agents was again left to the discretion of the anesthesiologist. After intubation, patients were mechanically ventilated with a Servo Siemens-Elema 900C ventilator. The ventilator used a high fresh gas flow rate (40 to 60 L/min), thus preventing rebreathing of gas by the patient. An in-line isoflurane vaporizer was attached to the inspiratory limb of the ventilator. A stable end-tidal CO₂ between 30 and 35 mm Hg was achieved in all cases before beginning O₂ measurements. Once again, a 10-min equilibration period prior to 10 consecutive 1-min measurements of O₂ was utilized. All patients were hemodynamically stable (systolic blood pressure [SBP] > 90 mm Hg and no greater than a 15% deviation in SBP from baseline reading) before beginning O₂ measurements. O₂ measurements in mechanically ventilated patients were completed before surgical incision.

Interval data such as age, height, weight, BMI, BSA, and sedative doses are expressed as mean ± SD and are compared with a two-tailed Student's t test. Categorical data in each group are compared by chi-square test or Fisher exact test when appropriate. Repeated-measures two-way analysis of variance (ANOVA) was used to compare O₂ measurements during spontaneous breathing and mechanical ventilation in obese and control groups. Student-Newman-Keuls test was used for evaluation of results found to be significant by repeated-measures ANOVA. O₂ measurements were also standardized to BMI and again subjected to repeated-measures ANOVA. All O₂ measurements are graphically represented as mean ± SD. All values were considered significant at p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

There were 18 morbidly obese and eight control patients enrolled in the study. Patients did not differ by age or sex but did differ by height, weight, BMI, and BSA. These data are summarized in Table 1. In the morbidly obese group, there were nine patients with hypertension, five with diabetes, five with obstructive sleep apnea, two with hyperlipidemia, and two with mild stable asthma (no pulmonary symptoms and a normal lung examination). The control group had one patient with diverticulitis and one with hypertension, diabetes, and hyperlipidemia. Neither group had any patients with chronic lung disease such as emphysema or interstitial lung disease. There were three smokers in the morbidly obese group and two smokers in the control group (p = 0.63). All three obese smokers had quit 10 yr ago or greater and had 15, 20, and 23 pack-year histories of smoking, respectively.

All patients in both obese and control groups achieved the level of preoperative sedation described in METHODS. Midazolam and fentanyl were the drugs used to achieve preoperative sedation. The doses of midazolam and fentanyl did not differ between obese and control groups. General anesthesia was induced with either propofol or thiopental, followed by a nondepolarizing neuromuscular blocking agent (pancuronium, cisatracurium, or rocuronium). Fentanyl and isoflurane were used for maintenance anesthesia in all cases. There were no differences in the doses of inhaled isoflurane used for general anesthesia when comparing obese and control groups. The anesthetic regimens are summarized in Table 2.

The overall changes in O₂ from spontaneous breathing to positive pressure ventilation were significantly different between obese and control groups (p < 0.0001 by two-way repeated-measures ANOVA). Baseline O₂ was higher in the obese patients compared with the control group (354.6 versus 221.4 ml/min; p = 0.0001 by Student-Newman-Keuls test; Figure 1). The change in O₂ from spontaneous breathing to positive pressure ventilation was highly significant in the morbidly obese patients (354.6 versus 297.2 ml/min; p = 0.0002 by Student-Newman-Keuls test). In contrast, the control group showed no significant change in O₂ from spontaneous breathing to positive pressure ventilation (221.4 versus 219.8 ml/min; p = 0.86 by Student-Newman-Keuls test). These results are illustrated in Figure 1.

View larger version (17K): [in this window] [in a new window]   Figure 1.   Mean (± SD) O2 in morbidly obese and control patientsspontaneous breathing versus positive pressure ventilation.

The changes in O₂/BMI from spontaneous breathing to positive pressure ventilation were significantly different when obese and control groups were compared by ANOVA (p < 0.0011). The baseline O₂/BMI was lower in the morbidly obese patients compared with the control group (6.4 versus 10.2 ml/ kg/m²; p = 0.00013 by Student-Newman-Keuls test). In the morbidly obese group, there was again a significant decrease in O₂/BMI from spontaneous breathing to mechanical ventilation (6.4 versus 5.4 ml/kg/m²; p = 0.00016 by Student-Newman-Keuls test). The control group showed no change in O₂/ BMI after mechanical ventilation (10.2 versus 10.1 ml/kg/m²; p = 0.75 by Student-Newman-Keuls test). These results are illustrated in Figure 2.

View larger version (19K): [in this window] [in a new window]   Figure 2.   Mean (± SD) O2/BMI in morbidly obese and control patientsspontaneous breathing versus positive pressure ventilation.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Morbid obesity may predispose patients to develop respiratory failure for a variety of reasons (10). Sharp and colleagues found obese patients to have reductions in total respiratory compliance and increases in total work of breathing (11). The accuracy of measurements performed in this study depended on a patient's ability to voluntarily relax respiratory muscles and allow a tank respirator to perform his or her respiratory work. The investigators acknowledged that measurements made during voluntary relaxation would require confirmation by comparison with measurements made after pharmacological paralysis. In addition, this study did not directly measure O₂. Other studies have found a disproportionate increase in energy expenditure with increases in E above baseline (6, 8). Kaufmann and colleagues hypothesized that their findings of increased oxygen cost of breathing with incremental increases in ventilation might be "extrapolated back to the resting level, then obese individuals [perhaps] have increased costs of breathing even at rest" (8).

We sought to test this hypothesis by directly measuring the change in O₂ from spontaneous breathing to mechanical ventilation while paralyzed. We have previously shown that intubated, mechanically ventilated patients have a significantly higher O₂ while awake than they do after adequate sedation (12). Potential increases in O₂ related to mechanical ventilation with an endotracheal tube in place in the present study were eliminated by assuring that these measurements were made under general anesthesia. In addition, preoperative anxiety, with its tendency to increase O₂, was minimized by assuring aggressive preoperative anxiolysis for all patients. Preoperative sedation was titrated based on the clinical response of each patient and the stated goals for depth of sedation were achieved in all patients studied. Induction of general anesthesia previously has been reported to result in a decrease in O₂ (13). Stevens and coworkers (13) noted a 17% reduction in O₂ from baseline values in patients anesthetized with isoflurane; however, baseline O₂ measurements were made without preoperative sedation and measurements under general anesthesia were made at a higher mean concentration of isoflurane (1.2%) than that used in our study. These two extremesanxiously awake and deeply anesthetizednot surprisingly, are clearly associated with two different states of oxygen consumption. Hirvonen and coworkers (14) likewise found a significant drop in O₂ from the awake, preoperative state to measurements done under general anesthesia. Although patients received oral diazepam (10 to 15 mg) for preoperative sedation in this study, their awake O₂ levels were higher than ours. Because they did not describe their patients' level of consciousness after preoperative sedation, it is difficult to compare the depth of preoperative sedation in our study with theirs. White and associates reported that the mere transition from wakefulness to normal sleep is associated with an 8.5% decrease in O₂ (16). Our preoperative sedation regimen sought to minimize changes in O₂ related to the transition from the awake state (where preoperative anxiety may even artificially elevate O₂) to the asleep, anesthetized state. If we could minimize the impact of general anesthesia on O₂, we hypothesized that the major variable impacting changes in O₂ in our study would be the energy dedicated to the work of breathing.

In contrast to the nonobese control patients, the morbidly obese group showed a striking decrease in O₂ during the transition from spontaneous breathing to positive pressure ventilation. We found a 16% reduction in mean O₂ in obese patients, compared with a less than 1% reduction in mean O₂ in the control patients. Even if one accepts the estimate of 3% of O₂ dedicated for O_2RESP in normals (3, 4), obese patients exhibited a fivefold increase in O_2RESP/O₂. If one couples this very large baseline difference between obese and normal patients to the greater increase in O_2RESP in obese patients with increasing E (6), it is likely that for a given metabolic or respiratory insult (e.g., the metabolic acidosis of sepsis or an acute pneumonia), morbid obesity would be a substantial risk factor for cardiopulmonary failure. This is particularly so in view of the inefficiency in respiratory muscle function seen both during normal breathing (4) as well as in acute respiratory failure (17). Robertson and colleagues found respiratory muscle efficiency in dogs to be lower than that measured for other skeletal muscles at all levels of inspiratory resistance (4). Manthous and coworkers found that intubated, critically ill patients had a 20% higher O₂ during spontaneous breathing with continuous positive airway pressure (CPAP) compared with full ventilatory support (assist-control) after muscle relaxation (17).

The obese patients demonstrated a significantly lower O₂ when standardized by BMI. This observation has been previously reported (2, 18) and may be due to the lower blood flow and metabolic rate of adipose tissue compared with lean body tissue. Nevertheless, the lower O₂ standardized to body size did not ameliorate the detrimental impact of morbid obesity on O_2RESP.

There were no differences in doses of preoperative sedative given to obese and control patients. This is a potential weakness of the study, because lesser sedative drug doses in the morbidly obese group (on a per kilogram basis) could conceivably lead to lesser preoperative sedation and therefore a higher baseline O₂. We believe this is unlikely because preoperative sedation was titrated to a clinical endpoint in all patients by anesthesiologists not involved in the study and blinded to O₂ measurements. It is well recognized that there is substantial interindividual variation in response to drugs used for preoperative sedation (19). Previous pharmacokinetic studies of midazolam in morbidly obese patients have found an increased volume of distribution and elimination half-life for these drugs compared with normal control subjects (20). However, such pharmacokinetic studies do not allow prediction of clinical onset of these drugs after initial intravenous administration. Because our endpoint for preoperative sedation was successfully reached in all patients, we believe an inadequate level of sedation biased toward the morbidly obese group is unlikely.

Because we did not monitor our patients with right heart catheterization, a remotely possible pitfall is our lack of certainty regarding systemic oxygen delivery. Since critical reductions in oxygen delivery are associated with decreases in O₂, our decreases in O₂ may have been, in part, due to oxygen delivery falling below this critical inflection point. This seems very unlikely, however, given that our patients showed no evidence of hemodynamic instability while our O₂ measurements were being made. In addition, others have shown that even in patients with known cardiovascular disease, induction of general anesthesia is not associated with a decrease in oxygen delivery below the critical delivery point (21).

In conclusion, we have demonstrated that morbid obesity is associated with a substantial increase in O_2RESP during quiet breathing when compared with normal control patients. The increase in energy expenditure which morbid obesity produces suggests a limited ventilatory reserve which may predispose such patients to respiratory failure during acute pulmonary or systemic illnesses.