Aging is associated with a reduction in anesthetic requirements. The effects of age on the electroencephalographic response to inhalational anesthesia have not been well documented. The objective of the present study was to determine the influence of age on hypnotic requirement and electroencephalographic derivatives such as bispectral index and 95% spectral edge frequency associated with sedation induced by sevoflurane.
Ninety-six patients were randomly allocated into one of three age groups A, B, and C, ranging in age from 18-39 yr, 40-64 yr, and 65-85 yr, respectively. Patients in each group were sedated with sevoflurane at two predetermined concentrations ranging between 0.45% and 0.85%. The relationship between sevoflurane concentration and response to a verbal command, as well as the relationships between response and bispectral index and 95% spectral edge frequency, was determined.
Multiple regression analysis showed that end-tidal sevoflurane concentration and age significantly affected both bispectral index and 95% spectral edge frequency. ED50 values of sevoflurane concentration for loss of consciousness, defined as no response to verbal command, were different between groups A and C: 0.72 (95% confidence interval: 0.68-0.75) versus 0.59 (95% confidence interval: 0.56-0.62). However, the same effective values of bispectral index and 95% spectral edge frequency at this same clinical end point did not differ.
Increasing age reduced sevoflurane requirements to suppress responses to a verbal command but did not change bispectral index and 95% spectral edge frequency associated with this end point, and in a population with a wide age range, bispectral index would predict depth of sedation better than end-tidal sevoflurane concentration.
AGE is one of the many factors known to influence the minimum alveolar concentration (MAC) that prevents movement in response to a noxious stimulus in 50% of a population, or that suppresses response to a verbal command (MACawake). 1–4An inverse relationship between age and anesthetic or hypnotic requirement is reported for sevoflurane and isoflurane. 4,5The electroencephalogram (EEG) is known to change with the depth of sedation and general anesthesia level. Recent studies demonstrated that real-time processed EEG data, such as bispectral index (BIS) or 95% spectral edge frequency (95% SEF), correlate well with hypnotic and sedative end points such as sedation level, loss of consciousness, lack of awareness, and memory. 6,7The BIS and 95% SEF values at which 50% of patients are adequately sedated have been determined in many studies, and this information is clinically valuable. Although BIS is reported to be independent of the anesthetics used, 7the influence of patient age on BIS or 95% SEF has not been well documented.
The objective of the present study was to determine the influence of age on hypnotic requirement, BIS, and 95% SEF associated with sedation induced by sevoflurane.
Material and Methods
After approval for this study had been obtained from the ethics committee of Hamamatsu University Hospital, informed consent to participate was acquired from all patients. The study group included 96 patients having American Society of Anesthesiologists physical status of I or II who were scheduled for elective surgery. Exclusion criteria were a history of cardiac, pulmonary, or renal disease; a history of esophageal reflux or hiatal hernia; drug or alcohol abuse; significant obesity (body mass index > 30); and contraindication for an inhalational induction.
Patients were assigned to one of three age groups, each having 32 subjects. Groups A, B, and C ranged in age from 18–39 yr, 40–64 yr, and 65–85 yr, respectively. Patients fasted for at least 8 h before surgery and received no premedicant drugs.
Electroencephalographic electrodes (Zipprep, Aspect Medical Systems, Natick, MA) were placed in the following configuration: bipolar frontomastoid montage (Fp1-A1 and Fp2-A2 in the International 10-20 System of electrode placement). The impedance of each electrode was less than 2 kΩ. EEG parameters including BIS (software version 3.2) and 95% SEF were recorded continuously using an Aspect A1,000 EEG monitor. Data were averaged online from the combined two leads because similar results were obtained from two channels. The BIS and 95% SEF values were calculated by averaging the values recorded for these variables during the 45-s interval immediately before assessment. To minimize artifacts, patients were instructed not to open their eyes, talk, or move during the EEG recording. Four artifact-detection schemes (slew rate, suppression, motion, and high frequency) were enabled. If these artifacts were detected, EEG parameters were excluded from contributing to the mean BIS or SEF. Serial output files consisting of processed EEG parameters were collected on a personal computer.
All patients breathed through a face mask connected to a semiclosed anesthetic circuit. To prevent contamination of end-tidal samples with inspired gas, the dead space was augmented at the sampling port using an extension tube. Gas was drawn continuously from the sampling port, which was located between the face mask and the dead space. The concentrations of carbon dioxide, sevoflurane, and oxygen were measured continuously by means of an infrared anesthetic gas analyzer (Capnomac, Datex, Helsinki, Finland), which was calibrated using a standard gas mixture before anesthesia for each patient. These concentrations were recorded every 10 s throughout the study period simultaneously with EEG parameters using a desktop computer (Macintosh Classic, Apple Computer, Tokyo, Japan).
Patients were sedated with sevoflurane and oxygen first during spontaneous ventilation, then during assisted ventilation if required, to keep the tidal volume high enough to measure the end-tidal anesthetic concentrations. End-tidal carbon dioxide concentration was kept between 35 and 45 mmHg during the study period. Inspired concentration of sevoflurane was adjusted to maintain the measured end-tidal concentration constant at a value set according to predetermined randomization scheme. After maintaining this target concentration for 15 min to allow blood:brain equilibration of sevoflurane, depth of sedation was assessed using the responsiveness component of the Observer’s Assessment of Alertness/Sedation (OAA/S) rating scale (table 1). 8This assessment procedure involves presentation of progressively more intense stimulation, ranging from a moderate speaking voice to physical shaking or moderate noxious stimuli (trapezius squeeze), until a response is observed. These stimuli were applied at 15-s intervals. The EEG parameters were recorded just before assessing sedation level. To calculate ED50values of sevoflurane concentration, BIS, and 95% SEF for loss of consciousness, we judged patients to be conscious if the OAA/S score was between 3 and 5 and unconscious if the OAA/S score was < 3. This assessment was performed twice per patient at different sevoflurane concentrations. The first concentration was selected randomly from among 0.50, 0.55, 0.60, and 0.65% for groups A and B and from among 0.45, 0.50, 0.55, and 0.60% for group C, and the second concentration was 0.2% higher than the first concentration.
To minimize interobserver variability, one investigator (T. K.) blinded to sevoflurane concentrations and EEG parameters performed all assessments.
Multiple linear regression, using a statistical software package (Statview II, Abacus Concept, Berkeley, CA), determined if end-tidal sevoflurane concentration, age, or gender affected BIS or 95% SEF.
The performance of hypnotic depth indicators such as sevoflurane concentration, BIS, and 95% SEF to predict depth of sedation was evaluated using prediction probability (Pk), which compares the performance of indicators having different units of measurement. The mathematical basis of Pk was described by Smith et al. 9A Pk value of 1 means that the predicting variable (e.g. , hypnotic depth indicator) always correctly predicts the value of the variable to be predicted (e.g. , true observed depth of sedation). A value of 0.5 means that the values of the indicator predict no better than a 50:50 chance (like flipping a fair coin). A value of Pk was computed for each group, and also for combined groups. The jackknife method was used to compute the standard error of the estimate, based on the assumption that the verbal commands were independent. A paired-data jackknife analysis was used to determine whether the Pk value of sevoflurane concentration differed from those of the EEG parameters within a group. To compare the performance of an indicator among the groups, a t test was performed using the jackknife results. For multiple comparisons, we applied the Bonferroni correction. Calculation of Pk was performed by using a custom spreadsheet macro.
We estimated ED50values of sevoflurane concentration, BIS, and 95% SEF for loss of consciousness using a logistic regression analysis described by Waud. 10The ED50values for the three age groups were compared by Waud’s technique. P < 0.05 was considered significant.
Characteristics of the 96 patients participating in this study are presented in table 2. Multiple regression analysis showed that end-tidal sevoflurane concentration and age significantly affected both BIS and 95% SEF. Both BIS and 95% SEF decreased in a dose-dependent manner with increasing end-tidal sevoflurane concentration (figs. 1 and 2).
Pk values for predicting depth of sedation are presented in table 3. In any age group, all indicators predicted depth of sedation significantly better than a 50:50 chance. No significant difference in Pk was observed between sevoflurane concentration and BIS, and Pks for BIS and sevoflurane concentration differed significantly from those for 95% SEF (P < 0.05). In the combined group, the Pk values (based on all 192 assessments) were 0.883 ± 0.013 (SE) for sevoflurane concentration, 0.926 ± 0.010 for BIS, and 0.831 ± 0.017 for 95% SEF. Pks for BIS and sevoflurane concentration differed significantly from those for 95% SEF, and Pk for BIS was different from that for sevoflurane concentration (P < 0.05). There were no differences in Pk of any indicator between groups.
The relationships between sevoflurane concentration, BIS, and 95% SEF and the probability of loss of consciousness for the three age groups are presented in figures 3–5. ED50values of sevoflurane (equivalent to MACawake), BIS, and 95% SEF for the three groups and the combined group are listed table 4. There was a significant difference in the ED50value of sevoflurane between group A and B; no difference was observed in the ED50values of BIS and 95% SEF.
The EEG patterns are known to change in relation to the depth of sedation 7and level of general anesthesia. 11,12In the present study, BIS was related linearly to sevoflurane concentration over the entire study range in all age groups. This finding is consistent with previous studies on sevoflurane and isoflurane. 6,7BIS seemed to decreased linearly with increasing anesthetic concentrations in a sedative range, with inhalational as well as intravenous anesthetics. 7,1395% SEF was also related linearly to sevoflurane concentration over the entire study range in all age groups. Previous studies demonstrated that quantitative EEG parameters such as 95% SEF varied predictably with increasing end-tidal sevoflurane concentrations, and that the EEG response to increasing sevoflurane was biphasic: An initial increase of EEG parameters from the awake state was followed by EEG slowing and a decrease in the EEG parameters. 6,14In the present study, we did not include EEG data obtained at end-tidal concentrations less than 0.4%, at which the EEG parameter would initially increase with increasing sevoflurane concentrations. Our findings on the linear relationship between sevoflurane concentration and EEG parameters agree with those of these previous studies.
Altered sensitivity to anesthetics in elderly patients has been widely reported for both inhalational and intravenous anesthesia. 2,3,15–20For intravenous anesthetics, these changes in drug action are caused by age-related alterations in drug disposition, end-organ sensitivity, or a combination of both. Age-related changes in drug disposition of intravenous anesthetics have been shown in many studies. 16,17,20,21There were no age-related changes in brain responsiveness or pharmacodynamics if the EEG parameter is used as a measure of thiopental effect. 16,17In contrast, increasing pharmacodynamic sensitivity to propofol in elderly patients was reported. 22The effect of age on pharmacodynamics may be dependent on the intravenous drug used and may be different between intravenous and inhalational anesthetics.
The importance of age on the EEG response to anesthesia has not been well documented for inhalational anesthetics. EEG changes in patients in response to different concentrations of isoflurane or halothane have been described, but the relationship of age to these responses has not been evaluated. 23–25We found that age was a significant independent variable in the analysis of BIS and 95% SEF during sedation induced by sevoflurane. A previous study showed that the proportion of time in EEG burst suppression was significantly greater in elderly patients than that observed in younger controls at 1.7% end-tidal isoflurane concentration, and that the frequency of EEG burst-suppression events was greater in elderly subjects. 26This demonstrates a discrete alteration relative to age in the central nervous system sensitivity to isoflurane. Another study showed that isoflurane and desflurane produced a similar increase in EEG burst suppression with increasing anesthetic concentration in neurosurgical patients, and that the EEG burst suppression was affected by the age of the patient. This is consistent with our findings, although the anesthetic concentrations used in the previous studies were much higher than those used in this one. Although it is clear that age affects MAC requirements during inhalational anesthesia, 1–3the relationship between EEG and MAC is inconclusive. EEG derivatives do not necessarily predict depth of anesthesia as defined by the response to surgical incision. 6,23This may be because of independent sites of action of anesthetics on EEG and MAC within the central nervous system. 27–29An increased brain sensitivity (as determined by EEG changes) in the elderly has been reported in association with fentanyl, alfentanil, and remifentanil. 15,20In contrast, other studies have demonstrated that age has no effect on the alfentanil plasma concentration required for suppressing responses to noxious stimuli. 18,30,31It is possible that changes in potency reflected in the EEG may not reflect changes in potency for some clinical end points.
The influence of aging on pharmacokinetics is an important factor in determining anesthetic requirements. We maintained target concentrations for 15 min to allow blood:brain equilibration of sevoflurane, which is probably long enough, 32,33although the influence of aging on blood:brain equilibration of sevoflurane has been unknown. BIS values were indeed stable a few minutes before depth of sedation was assessed.
Our previous study demonstrated MACawakedecreases at a rate similar to the rate of decreases in MAC relative to increasing age; therefore, the ratios to MAC can be considered fairly constant and are 0.34 for both sevoflurane and isoflurane. 4Although our finding of a decrease in MACawakein the elderly is consistent with that in the previous study, the rate of decrease in MACawakerelative to increasing age seems smaller. In other words, the decrease in MACawakeobtained in the elderly in the present study may be greater than that predicted with the regression line obtained in the previous study. However, the prediction of MACawakefor patients older than 65 yr may be overextended, given that the previous study did not include such patients. Several explanations for the discrepancy are possible. In the previous study, MACawakewas measured after surgery and the laryngeal mask airway was used, but in the present study MACawakewas measured before surgery and an anesthesia face mask was used. MACawakemeasured after surgery may be affected by the severity of postoperative pain. Severe postoperative pain would be a stronger stimuli with greater potential to awaken patients, resulting in a higher MACawakevalue. Relative to advancing age, the decrease in MACawakeobtained postoperatively would be greater than that in MACawakefree from postoperative pain, because pain thresholds in both somatic and visceral stimuli increase with advancing age. 34,35The influence of postoperative pain on MACawakemight be different between younger and elderly patients in the previous study.
Some observations suggest that motor responses to a noxious stimulus may be primarily mediated by subcortical structures including the spinal cord, at least in lower animals. 27–29In contrast, purposeful responsiveness to a verbal command apparently presupposes intact cortex function. These observations reveal that different sites of action determine MAC and MACawake. Therefore, a drug does not necessarily produce the same effect on both MAC and MACawake;i.e. , fentanyl reduced sevoflurane requirements for both loss of consciousness and surgical incision, but the reduction modes were not comparable at the two end points, 36and a single parameter that can predict responses at one end point does not necessarily predict responses at the other end point. 6Similarly, it is possible that the effect of the physiologic status of the patient, including variables such as age, does not produce the same change in both MAC and MACawake.
The finding that aging reduced MAC more than MACawakesuggests that, if physicians adjusted sevoflurane concentration using the movement/no movement end point (MAC end point) to provide adequate anesthesia, elderly patients would be awake (MACawakeend point) earlier on emergency from anesthesia than younger patients. However, this may not be the case with clinical anesthesia, because the end points used in anesthetizing surgical patients are usually hemodynamic parameters such as arterial blood pressure and heart rate, and the influence of age on MAC and blocking adrenergic responses remains unclear.
In contrast to the dependence of MACawakeon age with regard to application of sevoflurane, both BIS and 95% SEF values at which responses to loud verbal commands were suppressed in 50% of patients were not affected by the age of the patient. The influence of age on the efficacy of the EEG parameter to assess depth of sedation or anesthesia has not been well documented. We found no difference in Pk between the combined group and any age group. However, the Pk value for sevoflurane in predicting the depth of sedation did not differ from that for BIS in any age group, though the Pk value for sevoflurane significantly differed from that for BIS in the combined group. Further, MACawakevalue was significantly affected by age. We calculated Pk of the EEG parameters for predicting depth of sedation expressed on an OAA/S rating scale having six levels, not only for predicting the MACawakeend point. This means the EEG parameters are independent of age in predicting depth of sedation ranging from deep to light.
In conclusion, we found that (1) aging significantly affected the reduction in BIS and 95% SEF induced by sevoflurane; (2) during sevoflurane, aging reduced MACawakebut did not change BIS and 95% SEF associated with response to a verbal command; and (3) in a population with a wide age range, BIS would predict depth of sedation better than end-tidal sevoflurane concentration.
The authors thank the staff at Hamamatsu University Hospital for their support of this work, and Dr. Shibutani at New York Medical College for his encouragement and numerous suggestions.