Involuntary Muscle Contraction to Assess Nutritional Status
James S. Hayes1
Standard nutritional assessment techniques lack the sensitivity and specificity to identify early malnutrition and monitor short-term improvement from nutritional therapies. To improve on the available methods for nutrition assessment, Ross Products Division, Abbott Laboratories, has undertaken a research project with the goal of developing a device to study skeletal muscle function as an indication of nutritional status. The study of skeletal muscle as an indication of nutritional status was begun because of the well-known effects of malnutrition on muscle mass. Skeletal muscle function can be studied by the use of voluntary contractions (handgrip strength) or involuntary contractions (muscle contraction in response to an electrical stimulus).
Muscle function as measured by handgrip dynamometry has been shown to predict postoperative complications with a good degree of clinical accuracy (Kalfarentzos et al., 1989; Klidjian et al., 1980; Webb et al., 1989). Unfortu-
nately, the results of handgrip dynamometry are subject to patient effort and therefore may not be as sensitive and reliable as hoped.
The force, relaxation, and endurance characteristics of the adductor pollicis muscle in response to ulnar nerve stimulation may be a more reliable means of measuring muscle function than the measurement of voluntary muscle function. Since an electrical wave form with a standard frequency, strength, and duration is used to stimulate the ulnar nerve, the possibility that patient effort will confound the result is minimized. This technique has shown a diminution in muscle function in humans following both brief and prolonged periods of starvation (Russell et al., 1983a, b). In addition, refeeding has been shown to improve skeletal muscle function prior to appreciable changes in nitrogen status and muscle bulk (Brough et al., 1986). This improvement may occur as quickly as 4 days after the institution of nutrition support (Christie and Hill, 1990).
The basic technique that is used to study skeletal muscle function was developed in 1954 by Metron. Previous studies of ulnar nerve stimulation to assess muscle function have used strip chart recording devices to record results. The nature of this form of data collection may impair the sensitivity and possibly the specificity of the technique. This technique has been updated by the addition of a computer for collection and data analysis.
DESCRIPTION OF THE CURRENT MUSCLE FUNCTION ANALYSIS DEVICE
The current muscle function analysis (MFA) device uses a custom base plate to position the arm and force transducer, a stimulus generator with a constant current output and patient electrical isolation, and a standard IBM-compatible computer with MFA software installed. The primary factor responsible for electrical stimulation is current. However, the amount of voltage needed to produce this current is a function of the impedance presented to the source by the stimulating electrodes and the surrounding tissue. Two types of sources can be used to provide electrical stimulation, a constant voltage source or a constant current source. With a constant voltage source, the voltage waveform is sent to the electrodes, and the current waveform is dependent on the load impedance. The current value is dependent on the impedance of the tissue: the higher the impedance, the lower the current at a given voltage. A constant current source offers the advantage of being able to provide currents more independent of tissue and electrode impedance.
Figure 23-1 is a block diagram of the MFA device that illustrates the main components. The stimulus generator is a Grass Model S44 solid-state, square-wave stimulator. The stimulus pluse width, frequency, and intensity are controlled here. The isolation unit is provided to isolate the stimulus from the ground, reducing stimulus artifact, to provide a constant current for stimulation and to limit the maximum current that can be delivered to the patient to 15 mA. The patient interface consists of the base-plate assembly and force transducer.
The force transducer provides a voltage output proportional to thumb tension. These values are sent to the computer for analysis and storage.
The base-plate assembly consists of a fiberglass-reinforced polyester plate with foam armrests and an adjustable, pivoting mount for the force transducer. There are two armrests for ambidextrous use. The armrests are designed to position the arm and hand in a palm-up position with the wrist flexed and extended. This position facilitates the search for and location of the ulnar nerve. In addition, this position aligns the thumb with the force transducer.
The force transducer is a 25-lb, tension-load cell. It is mounted on a vertical cylinder that slides over a post. The post is mounted to a slide mechanism that is controlled by a crank at the front of the base plate. The cylinder mount can swivel about the post to align itself properly in the natural plane of the force being applied. By turning the crank and moving the post and transducer assembly, the static tension applied to the load cell and thumb can be adjusted for optimal performance. When the tension is adjusted properly, the post can be locked in place by tightening the hand nut on top of the post.
The electrical signals from the force transducer are amplified and sent to the computer. The computer displays the force curve data, performs some simple computations, and displays the results with other pertinent data. The computer also controls the stimulus start and duration times by gating the stimulus generator on and off. Stimulus pulse width, frequency, and intensity are set on the stimulus generator and are not controlled by the computer at this time.
With the arm of the subject placed on the base plate, the thumb loop extending from the force transducer is placed over the thumb. The hand crank is then rotated, moving the force transducer and applying a static force to the load cell. When the force is approximately 10 newtons, the post is locked into position. The electrodes from the stimulus generator are placed near the ulnar nerve, either at the wrist or elbow. A software routine that turns the stimulus on and off is initiated. This aids in locating the ulnar nerve but does not collect data. The frequency of the electrical stimulus is 1 to 5 Hz in the locate mode but can be adjusted to whatever value desired. During this location time, the electrodes are moved and gently pressed into the skin until a maximum force is elicited from the adductor pollicis muscle, as indicated on the computer screen. The computer is then switched into the data collection mode; the test sequence is started; and data is collected.
During the data collection period, contraction of the adductor pollicis muscle is caused by electrical stimulation of the ulnar nerve. The nerve is stimulated at frequencies of 10, 30, and 50 Hz at a current of up to 15 mA. The pulse duration is typically 500 µs. All of these parameters (frequency, intensity, and duration) can be varied as needed.
Contraction-relaxation curves at 10, 30, and 50 Hz are recorded by the computer. From these curves, a variety of descriptors of muscle function are obtained. These descriptors include the peak force generated at 10, 30, and 50 Hz (F10, F30, and F50); relaxation rate at 10, 30, and 50 Hz (RR10, RR30, and RR50), and the force frequency ratios of F10/F50 and F30/F50. A typical contraction-relaxation curve is shown in Figure 23-2.
Once the stimulus is removed, the muscles will relax. The rate of relaxation has been shown to differ between adequately nourished and malnourished individuals (Lopes et al., 1982). Relaxation rate is calculated at each frequency of stimulation by:
where F1 – F2 is the 10-µs period on the relaxation portion of the curve with the greatest force difference; Fp is the peak force; and FB is the baseline force. This calculation of relaxation rate results in a percentage of relaxation. In addition, the rate in newtons per second is calculated. The 10-µs period is based on the 100-Hz sampling rate of the analog-to-digital converter.
Force frequency ratios may be another indication of nutritional status. These are ratios of the force of contraction at a frequency not expected to produced maximal contraction (10 or 30 Hz) to the force of contraction at the frequency of maximal contraction (50 Hz). Force frequency ratios are calculated by:
where n is 10, 30, or 50 Hz; max is the maximum frequency; Fpn is the peak force at frequency n; Fmaxp is the peak force at the maximum frequency; FBn is the baseline force at frequency n; and FmaxB is the baseline force at the maximum frequency.
In addition to these descriptors, rise time, decay time, and area under the curve are calculated.
CURRENT CLINICAL TRIALS
There are a number of papers in the literature in which muscle function has been used as an indication of nutritional status in both normal subjects and various patient populations (Berkelhammer et al., 1985; Brough et al., 1986; Christie and Hill, 1990; Lopes et al., 1982; Russell et al., 1983a, b). The MFA device used in these previous studies consisted of a stimulus generator, base plate with force transducer, and strip chart recorder to record the muscle contraction-relaxation curves. A computer was not used to calculate the muscle function descriptors. The clinical trials that are discussed in this section use the device developed by Ross Products Division, Abbott Laboratories, described earlier.
All of these studies offer proof of the concept of muscle function as an indicator of nutritional status and verify previous results of other investigators. All studies are ongoing. Populations studied include normal subjects, HIV-AIDS patients, trauma patients, renal dialysis patients, and nutritionally compromised hospitalized patients.
The purpose of this study was to establish the contractile characteristics of skeletal muscle in normal, healthy volunteers during ulnar nerve stimulation of the adductor pollicis muscle. These results potentially would pave the way for the study of muscle function in a variety of disease states.
One-hundred-four normally nourished volunteers aged 18 to 90 were recruited for this study. Subjects were distributed evenly between four age categories; 18 to 35, 36 to 55, 56 to 75, and 76 years and older. Each age group included 13 males and 13 females. A normal nutrition status was determined by a nutrition screen, which included measurements of height, weight, triceps skinfolds, midarm muscle circumference, serum albumin, and hemoglobin. Nutritional status was determined by the criteria listed in Table 23-1.
The presence of any two or more of the six factors listed in Table 23-1 was considered to place the subject in the ''nutritional risk" category and therefore deemed the subject ineligible for the study. If a subject met only two criteria, these criteria had to be in two different categories for the subject to be ineligible for the study.
Each subject was tested on three separate occasions, with not more than 7 days between successive tests. Both hands were tested each time. In the 24-h period prior to each test, the subjects were asked to limit alcohol intake to the equivalent of two 12-oz beers. In addition, subjects were asked to refrain from strenuous physical exercise for 48 hours prior to each test. The effects of exercise on muscle function are not clear. However, it has been reported that moderate exercise does not alter the electrically evoked peak twitch torque (Sale et al., 1992), nor the contractile properties of muscle (Barnard et al., 1970). Disuse atrophy has been reported to result in faster relaxation (Simard et al., 1982), which is opposite of what is seen in malnutrition.
Results from this study indicate that the mean value of RR10 is similar to that reported in the literature. Some of the subjects in this study fell into the reported "abnormal" range. Other results included differences in:
RR10 when comparing males with females older than 75 years,
RR30 when comparing males with females older than 55 years,
RR50 when comparing males with females 18 to 35 and older than 55 years,
RR50 in males older than 75 years and males younger than 75 years, and
RR30 in females older than 56 years and females younger than 56 years.
The objective of this study is to determine the ability of the MFA device to detect and monitor the effects of malnutrition and refeeding. To date, 30 patients
TABLE 23-1 Criteria for Determining Nutritional Status
Category 1: Anthropometric Criteria
1. < 90% ideal body weight.
2. Midarm muscle circumference < 10th percentile and/or triceps skinfold < 10th percentile.
Category 2: Nutritional Criteria
1. Energy intake < 30–35 kcal/kg for women and 30–40 kcal/kg for men as determined from evaluation of 24-h dietary recall records.
2. Protein intake < 0.8 g protein per kg of body weight based on ideal body weight.
Category 3: Biochemical Criteria
1. Serum albumin < 3.5 g/dL.
2. Hemoglobin < 14 g/dL for men and 12 g/dL for women.
have been studied. Nutritional assessment measurements are made at baseline and after 3 months of nutritional intervention. Results to date indicate improved nutritional status as indicated by MFA at the 3-mo visit.
The objective of this study is to determine if a peptide-based protein is better than an all amino acid-based protein in affecting the survival and nutritional status of trauma patients. Patients are studied using MFA and a variety of other techniques at baseline (admission) and at 5, 10, 15, and 21 days. Initial results indicate that the MFA relaxation rate appears to follow total body protein.
Renal Dialysis Patients
The objective of this study is to determine if long-term nutritional support can improve outcome and survival in malnourished renal dialysis patients. Patients are evaluated at baseline and at 5 and 10 months. No results have been reported as of yet.
Nutritionally Compromised Patients
The objective of this study is to determine the ability of the MFA device to detect and monitor the effects of malnutrition. In addition, the practicality of this method as a manageable, noninvasive, bedside tool is being investigated. Fifteen patients have been studied to date. Subjective global assessment is used to clas-
sify nutrition status. Patients are studied at baseline and at 10 days. Anecdotal reports indicate improved MFA parameters after feeding.
The potential advantages of the MFA device include:
It may improve on currently available nutritional assessment techniques.
It may detect early malnutrition because of its sensitivity to functional changes in feeding.
It is noninvasive. No blood draws are required for this device.
Results of the test are available immediately.
It is a low-cost procedure. Neither costly equipment nor dedicated personnel are required.
The disadvantages of the current device include:
The device is large and cumbersome.
There is some difficulty locating the ulnar nerve.
Once the nerve is located, there is some difficulty maintaining the location.
There may be some voluntary component to what is supposed to be an involuntary contraction. This may be due to the subject's anticipation of or overreaction to the stimulus.
There may be discomfort. Individual pain tolerances differ. Although the current delivered is less than 15 mA, some people find the stimulus uncomfortable.
The device is being redesigned to address the above disadvantages.
The conclusions drawn from this presentation are that prototype units are currently in use in clinical trials. The results to date are encouraging: there seems to be a correlation between muscle force descriptors and nutritional assessment parameters. The device is being redesigned to address the problems identified in the initial clinical trials.
It is hoped that this technology will offer a significant improvement over the currently available techniques because of the advantages listed above. Further clinical trials are needed to define the role of the MFA device in the assessment of nutritional status, and the trials are ongoing to validate its use in a variety of populations.
Barnard, R.J., V.R. Edgerton, and J.B. Peter 1970 Effect of exercise on skeletal muscle. II. Contractile properties. J. Appl. Physiol. 28:767–770.
Berkelhammer, C.H., L.A. Lelter, K.N. Jeejeebhoy, A.S. Detsky, D.G. Oreopoulos, P.R. Uldall, and J.P. Baker 1985 Skeletal muscle function in chronic renal failure: An index of nutritional status. Am. J. Clin. Nutr. 42:846–864.
Brough, W., G. Horne, A. Blout, M.H. Irving, and K.N. Jeejeebhoy 1986 Effects of nutrient intake, surgery, sepsis, and long-term administration of steroids on muscle function. Br. J. Med. 293:983–988.
Christie, P.M., and G.L. Hill 1990 Effect of intravenous nutrition on nutrition and function in acute attacks of inflammatory bowel diseas. Gastroenterology 99:730–736.
Kalfarentzos, F., J. Spiliotis, G. Velimezis, D. Dougenis, and J. Androulakis 1989 Comparison of forearm muscle dynamometry with nutritional prognostic index as a preoperative indicator in cancer patients. J. Parenter. Enteral. Nutr. 13:34–36.
Klidjian, A.M., K.J. Foster, R.M. Kammerling, A. Cooper, and S.J. Karran 1980 Relation of anthropometric and dynamometric variables to serious postoperative complications. Br. J. Med. 281:899–901.
Lopes, J., D.M. Russell, J. Whitwell, and K.N. Jeejeebhoy 1982 Skeletal muscle function in malnutrition. Am. J. Clin. Nutr. 36:602–610.
Merton, P.A. 1954 Voluntary strength and fatigue. J. Physiol. 123:553–564.
Russell, D.M., L.A. Leiter, J. Whitwell, E.B. Marliss, and K.N. Jeejeebhoy 1983a Skeletal muscle function during hypocaloric diets and fasting: A comparison with standard nutritional assessment parameters. Am. J. Clin. Nutr. 37:133–138.
Russell, D.M., P.J. Prendergast, P.L. Darby, P.E. Garfinkel, J. Whitwell, and K.N. Jeejeebhoy 1983b A comparison between muscle function and body composition in anorexia nervosa: The effect of refeeding. Am. J. Clin. Nutr. 38:229–237.
Sale, D.G., J.E. Martin, and D.E. Moroz 1992 Hypertrophy without increased isometric strength after weight training. Eur. J. Appl. Physiol. 64:51–55.
Simard, C.P., S.A. Spector, and V.R. Edgerton 1982 Contractile properties of rat hind limb muscles immobilized at different lengths. Exp. Neurology 77:467–482.
Webb, A.R., L.A. Newman, M. Taylor, and J.B. Keogh 1989 Handgrip dynamometry as a predictor of postoperative complications: Reappraisal using age-standardized grip strengths. J. Parenter. Enteral. Nutr. 13:30–33.
JOHANNA DWYER: I have just two questions. In your first study where you were looking at the influence of sex and age, you made a comment about total body protein.
JAMES HAYES: That was in trauma [patients].
JOHANNA DWYER: How do you measure the total body protein?
JAMES HAYES: To be perfectly honest with you, I am not sure. If you want to know, I can review the protocol and let you know.
JOHANNA DWYER: You said there was no effect of age, is that right?
JAMES HAYES: As you grow older, the relaxation rates tend to decrease.
JOHANNA DWYER: Okay, they did decrease. But you were not able to associate it with the total body protein?
JAMES HAYES: No attempt was made to do that in that particular study.
GAIL BUTTERFIELD: Are you making any attempt to quantify the degree of malnutrition as evidenced by the change in …
JAMES HAYES: I think that one of the goals of or hopes for the device is that we will be able to quantify the degree of malnutrition rather than say that you are normally nourished or you are malnourished. That is a goal. At the current time, we are making no effort or attempt to do that. The attempt is to determine that the device works and we actually are getting results that indicate malnutrition or adequate nutrition.
ARTHUR ANDERSON: I think an important control to include in your study, especially since you made an observation that correlated muscle or body protein with delayed relaxation time, is to get thyroid function tests to correlate with that. That is a diagnostic feature of hyperthyroidism, and you would expect a low body protein or malnourished appearance in someone who is hyperthyroid.
JEFFERY ZACHWIEJA: Do you have any information on activity status and relaxation time?
JAMES HAYES: What we recommend for the studies right now is that there is no vigorous upper-body exercise for 24 hours prior to the study. I have a feeling that fatigue will affect the results, but I have not studied that. We are trying to
eliminate that as a possibility. In the patient populations, they are normally not really active.
WM. CAMERON CHUMLEA: Have you thought about using some other body limb? I know the hand is easy because it is very accessible. It has been used for grip strength or for arm circumference, and those do have relationships to body composition, nutritional status, and muscle strength.
However, they do not appear to be as sensitive to the early stages and that is why a lot of the measures of those have now shifted to quadriceps strength or calf circumference because they do tend to be more sensitive to the early losses of nutrition.
JAMES HAYES: Yes, we considered any body limb to which we can attach an accelerometer or transducer that can respond to motions. One of the reasons to choose the hand is that even nonmobile patients may be using their upper bodies to transfer from bed to chair, so this muscle would be less likely to become atrophied. That is why we used that. But we could just as well use any other muscle group.
DOUGLAS WILMORE: If you think about the application of this to military personnel, you must realize that it is a functional test and that acute fluid and electrolyte changes will change muscle function, and because that is part of many states of malnutrition, this is not a malnutrition test, but it is a functional test. It may have some real advantages in that if you see muscle dysfunction, it may act as a red light as to whether you want to use personnel or not. But you do not know whether the results you see are due to electrolyte problems, you do not know whether it is salt restriction, you do not know whether it is low glucose, whether it is glycogen depletion or protein depletion.
JAMES HAYES: That is absolutely right, but one of the things, too, that we are finding is that the test is a good functional indicator of muscle function, which also may indicate nutritional status. Obviously, those things you mentioned may have an effect and we need to study those.
One of the things we wanted to do was get involved in a Ranger study so that we would able to follow those and other possible factors and have some data to support or to be more conclusive on that.
DOUGLAS WILMORE: It is just like phase angle [in bioelectrical impedance analysis], which may indicate a number of things—off or on, yes or no. This may provide another sort of measure.