Eur J Appl Physiol (2009) 106:583–588 DOI 10.1007/s00421-009-1054-1
O R I G I N A L A R T I CL E
EVects of spinal anesthesia on resting metabolic rate and quadriceps mechanomyography William Paul McKay · Brendan Lett · Philip D. Chilibeck · Brian L. Daku
Accepted: 25 March 2009 / Published online: 9 April 2009 © Springer-Verlag 2009
Abstract Previous work by our group has shown by mechanomyography (MMG) that resting muscle is mechanically active. Ten patients having spinal anesthesia for surgery, which paralyses muscle below the waist, were studied to help determine whether resting-muscle mechanical activity plays a signiWcant role in resting metabolism, and to further determine if the phenomenon is neurally mediated. Resting metabolic rate (RMR) by indirect calorimetry, and mid-anterior thigh MMG by accelerometer, were recorded before and during spinal anesthesia. Spinal anesthesia produced a 25% decrease in oxygen uptake (mean § standard deviation: pre-spinal 228 § 76; during spinal 171 § 67 ml min¡1; P < 0.001) and 37% decrease in mean absolute MMG signal amplitude (pre-spinal-anesthetic 10.6 § 3.9; during spinal: 6.7 § 3.5 mm s¡2; P < 0.001). Decreased oxygen uptake in individuals correlated with decreased resting-muscle mechanical activity (R = 0.624;
W. P. McKay · B. Lett Department of Anesthesia, University of Saskatchewan, Saskatoon, SK, Canada P. D. Chilibeck Department of Kinesiology, University of Saskatchewan, Saskatoon, SK, Canada B. L. Daku Department of Electrical Engineering, University of Saskatchewan, Saskatoon, SK, Canada W. P. McKay (&) Royal University, 103 Hospital Dr., Saskatoon, SK S7N 0W8, Canada e-mail:
[email protected];
[email protected]
P = 0.05). Paralysis of muscle below the waist reduced RMR and resting-muscle mechanical activity. Keywords Mechanomyography · Spinal anesthesia · Resting metabolic rate · Skeletal muscle · Human
Introduction Contracting muscle produces low frequency vibrations, known as mechanomyographic (MMG) activity, detectable at the skin overlying the muscle, and in a direction normal to the long axis of the muscle (Orizio 1993; Oster and JaVe 1980). Exercising muscle results in increased MMG activity and increased oxygen uptake (Stout 1997). Resting muscle also produces MMG activity, and resting-muscle MMG activity is increased after exercise and is associated with the increase in oxygen uptake (VO2) after exercise (McKay et al. 1998, 2004, 2006, 2007). Post-exercise MMG activity may therefore contribute to excess post-exercise oxygen consumption, and elevated metabolic rate. The purpose of the current study was to determine whether a reduction in MMG activity by paralyzing muscle would correlate with a reduction in resting metabolic rate (RMR). RMR in the absence of muscular work is considered to arise from basal cellular metabolism of all body cells, the work of the viscera, and the work of breathing. Resting-muscle MMG activity decreases with general anesthesia, and disappears with muscle paralysis by pharmacologic neuromuscular blockade (NMB) (McKay et al. 1998). The diVerence between RMR before and after paralyzing the muscles would enable calculation of RMR due to resting-muscle mechanical activity. In designing an experiment to test this model, NMB could be used to paralyze all of the skeletal muscles. NMB is used frequently in anesthesia
123
584
practice, but only in anesthetised patients. It is very distressing to be totally paralyzed and conscious. For the purpose of designing this experiment, we used spinal anesthesia, which produces a patient who is conscious and comfortable with all the muscle paralyzed below about the tenth thoracic spinal level (T10). This includes all lower limb muscles and muscles attached to the pelvis as well as the lower abdominal and lumbar spinal muscles. This comprises the majority of the body’s muscle mass; therefore, paralyzing these muscles provides a good model for studying the contribution of MMG to RMR. If resting-muscle mechanical activity contributes signiWcantly to RMR, the phenomenon is of considerable interest in basic research on exercise and obesity. We report an experiment to measure the RMR by indirect calorimetry before and during spinal anesthesia in order to determine whether the decrease in muscle mechanical activity is proportional to the decrease in RMR. This can be used to determine if a signiWcant portion of resting metabolic activity can be attributed to resting-muscle mechanical activity. Our hypothesis is that spinal anesthesia will decrease resting-muscle mechanical activity in myotomes aVected by the anesthetic and that this reduction will correlate with the reduction in RMR.
Methods Subjects With University of Saskatchewan Research Ethics Board approval, we recruited ten healthy ASA classes I and II (Mak et al. 2002) adult patients age 18–60 years who were having a spinal anesthetic for surgery. Signed informed consent was sought in the surgeon’s oYce at the time of arranging surgery, in Pre Admission Clinic, and in Day Surgery. The following patients were excluded from the study those who are unable to give informed consent, those with thyroid dysfunction, with fever or any systemic illness, those who chose general rather than spinal anesthesia, and those who might feel claustrophobic wearing the metabolic mask. Seven females and three males age (mean and standard deviation) 55 § 8.6 years, 166 § 11 cm tall, and weighing 84 § 16 kg were recruited. Nine subjects had orthopedic joint replacement procedures, and one had a general surgical procedure. Mean spinal level was Thoracic (T) 10 § 4 segments, with a range of T4 to Lumbar (L) 4. There was one missing value for post-spinal temperature. Equipment Heart rate and blood pressure were obtained on admission and in the Post Anesthetic Care Unit (PACU) by trained
123
Eur J Appl Physiol (2009) 106:583–588
nursing staV using an automatic non-invasive blood pressure cuV and Wnger plethysmograph. On admission a Critikon Dinamap Vital Signs Monitor model 8100 (GE Healthcare, Chalfont St. Giles, United Kingdom) was used, while in PACU a Philips Planar System model H80 monitor with a Philips 12XL model M3001A vital signs module (Koninklijke Philips Electronics N.V., Eindhoven, The Netherlands) was used. Temperature was taken sublingually using a digital clinical thermometer, Welch Allyn model 678 (Welch Allyn Inc, San Diego CA, USA). These devices are factory calibrated and Canadian Standards Association (CSA) approved for use on hospital patients. Electrocardiogram (ECG) standard monitoring leads (Medi-Trace, Kendall-LTP, Chicopee, MA, USA) were attached to enable lead-II monitoring (Spacelabs model #414 Monitor, Spacelabs, Spacelabs Healthcare, Issaquah, WA, USA). The ECG signal was used to remove the artefacts due to heart rate from the MMG signal as described in detail elsewhere (McKay et al. 2004). MMG recordings were made with an accelerometer (Bruel & Kjaer model #4381 accelerometer; 43 g, 2£2 cm; powered by a B&K model #2635 charge ampliWer; Brüel & Kjær Sound & Vibration Measurement A/S, Nærum, Denmark) that was calibrated with a Bruel & Kjaer model #4294 calibration exciter. The charge ampliWer provides pre-Wltering of the MMG signal at 2–100 Hz. The accelerometer was taped over the mid-quadriceps muscle belly with 3-M double-sided hypoallergenic tape, halfway between the anterior superior iliac spine and the upper pole of the patella. This is a single-axis accelerometer mounted with the active axis normal to the skin surface. Marking with a waterproof pen ensured the same sensor position after spinal anesthesia. A recording of simultaneous MMG and ECG of 2 min duration was made before and during the spinal anesthetic concurrent with the RMR measurements. ECG and MMG output was sampled at 1,000 Hz using a National Instruments DAQPad-6020E (National Instruments Corporation, Austin, TX, USA) attached to a laptop computer. Normal ECG signal content is below 250 Hz in children, and 150 Hz in adults (Rijnbeek et al. 2001), while MMG signal content is below 60 Hz (Beck 2007). Thus, the digitisation rate is well above the critical Nyquist frequency for both signals. The digitised signals were analyzed oV-line on a PC desktop computer, using tested signal analysis programs developed in our lab and described in detail elsewhere (McKay et al. 2004). MMG measurements of muscle mechanical activity (Fig. 1) are valid as they correlate well (r2 = 0.89) with torque reductions during fatiguing muscle contractions (Gobbo et al. 2006). The intra-class correlation coeYcient for repeated MMG measurements is 0.95 (Akataki et al. 1999). Metabolic measurements were made with a Cosmed K4B2 portable indirect calorimetry system (COSMED
Eur J Appl Physiol (2009) 106:583–588
Fig. 1 Ten-second segment of a typical MMG recording for same subject pre- and post-spinal anesthesia. X axis: time (s); Y axis on right applies to pre-spinal MMG amplitude (|amplitude| = 19.4 mm s¡2), Y axis on left to post-spinal MMG amplitude (|amplitude| = 12.1 mm s¡2)
S.r.l., Rome, Italy). This device was calibrated before each use according to the manufacturer’s instructions with a calibration gas with 4% CO2, 16% O2, and 80% N2 (Sensormedics, Yorba Linda, CA, USA). Flow/volume calibration of the turbine was done with a 3-l syringe provided by the manufacturer and according to manufacturer’s instructions. Metabolic measurements reported are oxygen utilization and carbon dioxide production in ml min¡1 (VO2, VCO2); energy equivalents (EE) in kcal per day and in metabolic equivalents, where one metabolic equivalent (MET) is deWned as the amount of oxygen consumed while sitting at rest; and RQ, the ratio of VCO2 to VO2. Despite short-term variability (Fig. 2), mean measurements of VO2 and VCO2 on the portable system correlate well to a standard laboratory metabolic cart, with r2 of 0.84 and 0.92, respectively, while the intra-class correlation coeYcients for repeated measures on the portable system range from 0.7 to 0.9 for ventilation, VO2, and VCO2 (DuYeld et al. 2004). Measurements of VO2 and VCO2 on the portable system have been validated by comparison to the same measurements assessed by a validated mixing box system consisting of O2 and CO2 analyzers and pneumotachograph (Doyon et al. 2001). There are no diVerences in measurement between the two systems at low to high exercise work rates. Measurements Usual demographics (age, gender, height, weight) and vital signs (temperature, heart rate, breathing frequency, blood pressure) were recorded, as well as any relevant clinical data (diagnosis, medications, surgical procedure). Vital
585
Fig. 2 Typical VO2 recording breath by breath for one subject preand post-spinal anesthesia. X axis: time (s), Y axis VO2 (ml min¡1)
signs, obtained in standard fashion with the subjects supine, were taken from the medical charts as charted by the Admitting and Post Anesthetic Care attending nurses. ECG, MMG, and RMR were recorded concurrently for 2 min while the patient was in bed in the surgical admitting area and again after surgery in the PACU after relaxing for 20 min. Level of the eVect of the spinal anesthetic was tested by cold sensation to ice. In all cases, the environment was calm and quiet. MMG is reported as mean absolute acceleration (see McKay et al. 2004). The MMG recordings were made on mid-anterior mid-thigh, halfway between the anterior superior iliac spine and the upper pole of patella because it is consistent, horizontal in the supine subject, easy to locate, even in thickset or obese subjects, gives a strong MMG signal because it overlies a large muscle mass, and is paralyzed by spinal anesthesia. Statistics Statistical analysis was performed with Sigmastat for Windows version 3.11 (Systat Software, Inc. Chicago, IL, USA). There are no similar experiments on which we could base formal sample-size calculations. Our experience in this Weld showed that we could expect signiWcant results with ten subjects. Descriptive statistics are reported for demographics, vital signs, and clinical data (mean, SD, distribution). MMG and metabolic measurements before and during spinal anesthesia were compared by paired t test. P · 0.05 was considered statistically signiWcant. VO2, VCO2 and EE were treated as one family of values, and temperature, pulse, respirations, and mean blood pressure as a second family of values for purposes of applying a
123
586
Eur J Appl Physiol (2009) 106:583–588
correction for multiple comparisons. Bonferroni’s correction was applied. Linear regression was performed to examine the relationship between changes in mean VO2 and mean MMG amplitude.
Results Metabolic, MMG, and cardiorespiratory results are presented in Table 1 and Fig. 3. A signiWcant portion (25%) of RMR disappeared with spinal anesthesia, as did a signiWcant portion of the mean absolute amplitude of the MMG signal (37%). EVect size for VO2 was 0.75 and for MMG was 1.0. Power of the paired t test for both VO2 and MMG amplitude with alpha = 0.05 was 1.0. There was a rather large variability between subjects. Small sample-size accounts for some of this. Additional factors possibly producing metabolic variance are subject size, variability in true RMR per kg (due to variability in age and body fat %), and measurement variance from inaccuracies in the metabolic measuring instrument. Age ranges from 50 to 68 years, and weight from 66 to 115 kg. For the post-spinal measurements, variance is due also to spinal-height diVerences due both to diVerences in the maximum eVective height of the spinal anesthetic and diVerences in the rate of descent of the spinal (how rapidly the spinal “wears oV”). Large variance in spinalheight diVerences and diVerences in the rate of the spinal anesthetic wearing oV are normal for spinal anesthesia (Greene 1985). There was a signiWcant correlation between the individual changes in VO2 using linear regression of the ratio of post-spinal/pre-spinal VO2 on the ratio of post-spinal/prespinal MMG mean absolute amplitudes: VO2 ratio = 0.534 + (0.354 £ MMG ratio); R = 0.624, P = 0.05. Ninety-Wve
Fig. 3 Change in VO2 with spinal anesthesia for each subject. X axis: pre- and post-spinal anesthesia. X axis, phase of experiment; Y, axis: VO2 (ml min¡1)
percent conWdence intervals (CI) for the correlation coeYcients were 0.34–0.73 for the constant component, and 0.05–0.66 for the MMG ratio coeYcient.
Table 1 Metabolic and MMG Results Pre-spinala
Post-spinala
Percent change
P value
ConWdence intervals of the diVerence
VO2 (ml min¡1)
228 § 76
171 § 67
¡25
