The relative proportions of fiber types within muscle and the characteristics of these fiber types are important determinants of the surface electromyogram (SEMG) during fatigue. In this study, patients suffering from congenital myopathy characterized by a strong type I fiber predominance were studied. Six patients with 95- 100% type I fibers, 2 patients with 80% type I fibers, and 12 healthy volunteers participated in an ischemic, isometric, intermittent exercise test of m. quadriceps femoris at 80% MVC. Considering the results of the morphometric analysis of muscle biopsy specimen and of the anthropometric estimated muscle- bone volume, it was found that type I muscle fibers had a lower force generating capacity than type II fibers. The initial conduction velocity along the muscle fiber membrane (MFCV) was low in patients with 95-100% type I fibers. During the ischemic exercise test, the 95-100% type I fibers showed less fatigability than type II fibers, which was reflected by a nearby absent decrease of the muscle membrane excitability as measured by the MFCV, and only a slight increase of the SEMG amplitude compared with patients having 80% type I fibers and controls. The absence of a definite MFCV decrease was related to the nearby lacking lactate formation in 95-100% type I fibers. Key words: congenital myopathy type I muscle fiber predominance surface electromyography muscle fiber conduction velocity force MUSCLE & NERVE 14:829-837 1991
FATIGUE IN TYPE I FIBER PREDOMINANCE: A MUSCLE FORCE AND SURFACE EMG STUDY ON THE RELATIVE ROLE OF TYPE I AND TYPE II MUSCLE FIBERS WIM H.J.P. LINSSEN, MD, DICK F. STEGEMAN, PhD, Ed M.G. JOOSTEN, MD, PhD, ROB A. BINKHORST, PhD, MIEKE J.H. MERKS, BSc, HENK J. TER LAAK, PhD, and SERVAAS L.H. NOTERMANS, MD, PhD
T h e congenital myopathies (ie, central core disease [CCD] and nemaline rod myopathy [NRM]) are characterized by a generalized type I fiber predominance in all skeletal muscles, except for some muscles who are innervated by cranial nerves2,' The relative proportions of fiber types within From the Institutes of Neurology (Drs. Linssen, Stegeman, Joosten, Merks, ter Laak, and Notermans) and Physiology (Dr. Binkhorst). University of Nijmegen, The Netherlands. This investigation is part of the research program "Disorders of the neurornuscular system" of the University of Nijmegen. Acknowledgments: The authors thank Mrs. G. Steenbergen and Mrs. L. van Woerkom for their technical assistance and the laboratory examinations, and Mr. H. Wijnen and Mr. N. Dijkstra for preparing the figures. Address reprint requests to W.H.J.P. Linssen, MD, Institute of Neurology, University of Nijrnegen. P.O. Box 9101, 6500 HB Nijmegen, The Netherlands. Accepted for publication August 14, 1990.
CCC 0148-639X/91/090829-09 $04.00 0 1991 John Wiley & Sons, Inc.
Fatigue in Type I Predominance
muscle and the characteristics of these muscle fiber types are undoubtedly important determinants of the surface electromyogram (SEMG) during fatiguing exercise. Normally type I muscle fibers become recruited at low force levels. When the force level is increased, type I1 motor units become recruited in addition.34 The abovementioned congenital myopathies may serve as a model to study the type I fiber characteristics. Type I1 motor units show a fast action potential conduction velocity along the muscle fiber membrane (MFCV). This explains why the MFCV increases with increasing force levels.6229The larger MFCV is thought to be related to the larger diameter of type I1 fibers,6,8,'6although Sadoyama et al?' found the MFCV not positively correlated with the muscle fiber diameter. In their experiment, the MFCV is related to the cross-sectional area of type I1 fibers. During fatiguing exercise, the power density
MUSCLE & NERVE
September 1991
829
frequency spectrum (PDS) shifts to lower frequenc i e ~ . ' ~Milner-Brown '~ et al.24 showed that (unfatigued) type I1 fibers have a higher median frequency. Braakhekke et al.5 found the PDS shifting toward higher frequencies in a patient suffering from a carnitine deficiency during submaximal (30% of the maximal aerobic power) bicycle ergometric exercise over a 2-hour period. From these experiments it was suggested that, like the MFCV, the PDS and its shift during fatigue were related to the relative role of type I1 and type I fibers. Another aspect is whether type I muscle fibers generate more force than type I1 fibers. Some authors find no significant differences between the force-generating capacity of human type I and type I1 muscle fibers"; others find type I1 muscle fibers more powerful.35 Some studies2123'suggest that strength is correlated to the muscle crosssectional area irrespective of the fiber composition. Animal s t ~ d i e s ~ ~ have ' " ' ~ shown that type I1 fibers are capable of generating more force than type I fibers. T h e present study had two aims. By studying the above-described congenital myopathies, it became possible: (a) to determine the forcegenerating capacity, the MFCV and SEMG frequency spectrum parameters of muscle depending on the fiber type composition; and (b) to investigate the changes of the MFCV, the SEMG frequency spectrum and amplitude during anerobic fatiguing exercise, again depending on the fiber type composition and some metabolic parameters. MATERIALS AND METHODS Patients and Controls. six mean age 23.4 ? 7.6 years,
patients (5 females: mean height 1.58 0.08 m, mean weight 46.8 5 8.7 kg; 1 male: 13 years; height 1.40 m, weight 29 kg) with 95- 100% type I fiber predominance together with two patients (1 female: 15 years, 1.64 m, 40 kg; l male: 35 years, 1.74 m, 72 kg) with SO% type I muscle fibers in their muscle biopsy specimen performed isometric ischemic intermittent m. quadriceps femoris contractions at 80% maximal voluntary contraction (MVC), rate 30/min. Twelve healthy volunteers (8 females: mean age 30 k 6.4 years, mean height 1.71 2 0.03 m, mean weight 60 ? 6.8 kg; 4 males: mean age 36 ? 11.3 years, mean height 1.79 ? 0.16, mean weight 74.5 ? 10.2 kg) served as a control group. As all controls were without neuromuscular complaints, no muscle biopsies were taken. A normal mixed distribution with more than 45% type I1 muscle fibers within the m. quadriceps femoris was assumed." All pa-
830
Fatigue in Type I Predominance
*
tients and control subjects gave their informed consent. The Committee on Experiments in Humans of the Faculty of Medicine saw and approved the experimental protocol. (Table 1) Of the patients with 95100% type I fibers, 5 were suffered from central core disease (CCD), and 1 patient from adult autosoma1 dominant nemaline rod myopathy (NRM). Of the 2 patients with 80% type I fibers, one suffered from CCD. The other patient complained of exertional myalgia symptoms. In her biopsy specimen, no structural abnormalities were found except for a marked type I fiber predominance. In the congenital myopathies, the amount of structural abnormalities (cores and rods) and fat cells may vary per muscle group.' In most cases, the muscle biopsies were taken from the m. quadriceps femoris. Morphometric studies were carried out using SDH or NADH-TR stained sections. Area measurements of the muscle fibers and of the unstained parts (representing the areas containing cores or nemaline rods) of these fibers were performed by using a systematic random point sampling design. In this way, the mean fiber area, the relative area of cores or rods, the percentage of fibers containing cores, and the percentage of area of fat cells were calculated. The area measurements were adjusted to age at the time of biopsy. T o relate the physiological results to the gross anatomical structures of the limb, skinfold measurements (percentage fat)' l were carried out, and the mean anthropometric estimated upper arm muscle-bone area (MBA [cm2])" was determined in most subjects. T h e MBA measurements were standardized for the upper arm muscles and, therefore, offered only indirect information about the muscle volume of the m. quadriceps femoris. Biopsy Data.
Experimental Set-Up. T h e experimental protocol of the isometric ischemic m. quadriceps femoris test resembled the ischemic forearm exercise test (IFET) as described by Sinkeler et al.32 In short, all subjects were seated upright on a table with the upper leg supported and the lower leg hanging freely. The ankle was firmly attached to a strain gauge dynamometer (Fig. lA).3" The maximal voluntary contraction level (MVC) of the right m. quadriceps femoris was determined twice with 1 minute of rest in between. T h e amplified force signal was recorded on paper. Thereafter, all subjects relaxed for 10 minutes. During the test, intermittent contractions were
MUSCLE & NERVE
September 1991
~~~
~
~
~
~
Table 1. Morphometric studies 80% Type I
95-100% Type I
Sexlage (yr) Biopsy site Type lldisease (%)
Number of fibers counted Mean fiber* area (pm2) Mean area with coreslrods per fiber (%) Area with fat cells (%) Estimated reduction of contractile elements (%)
1
2
3
4
5
6
7
8
MI13* Erec. tr 100 CCD
Fl15 quadr 100 CCD
F/21t quadr. 96 CCD
Fl20
soleus 100 CCD
F126 quadr 100 CCD
Fl35 quadr 100 NRM
MI35 quadr 80 CCD
F/l5 quadr. 80 type 1
112 809 9
I34 4321 14 11 23
72 936 3 14 16
112 2208 21 11 30
7 3226 43 33 62
68 4601 29 13 38
126 8052 7
75 5398 0 0.5 0
51
55
1 7
‘Biopsy at 7 year of age. f8iopsy at 5 years of age. #Calculations are made by statisticaliy transformation of the muscle fiber dmmeter variabie ( d ) into a new variable (a), so that a = ~rd’/4.In adult controls we suppose” that d has a normal distribution of about 60 pm’ and a variance of 64 pm’. It follows that the calculated normal mean fiber area (a) will be 2878 t 757 pm’. In our laboratory, the mean fiber diameter in normal subjects is assumed to be 20 ? 3 pm at age I and 30 2 5 pm at age 5 . It follows a normal mean fiber area of 327 2 95 pm’ at age 1, and 726 2 237 pm’ at age 5 .
made at 80% of MVC under ischemic conditions. Ischemia was applied by inflation of a cuff above 200 mmHg pressure 5 seconds prior to the experiment. The cuff was wrapped around the m. quadriceps as proximal as possible. Intermittent contractions were made at a rate of 30 per minute until the 80% MVC force level could no longer be reached. This moment was defined as the failure point. Data analysis was performed over the performance time (PT)ending at the failure point.
A
Fatigue in Type I Predominance
Much care was taken in positioning the SEMG array parallel to the direction of the muscle fibers of the short head of right m. vastus medialis, distal to the motor point (Fig. 1A). We used a specially designed surface electrode array consisting of 4 gold-coated electrodes placed in-line. The electrode diameter was 2 mm and the interelectrode distance was 6 mm (Fig. IB). The skin was carefully abrased and cleaned with ethanol. N o electrode paste was used in order to avoid low-
FIGURE 1. (A) Subjects are seated and firmly attached to the quadriceps dynamometer. The right leg is horizontally fixed. The lower leg is hanging freely. The surface electrode array is positioned distally from the motor point on the short head of the m. vastus medialis. (B) Detail of the surface electrode array consisting of 5 electrodes. Four electrodes are used for deriving 2 bipolar SEMG signals.
MUSCLE & NERVE
September 1991
831
resistance bridges between the closely spaced electrodes. The experimental protocol, the way the SEMG signals were derived, and the data analysis have been described previously."' After amplification, band pass filtering, AID conversion, sampling, and storing on magnetic disk, the SEMG signals were compared by determining the cross-correlation function. The maximum cross-correlation should exceed 0.7. In 4 patients with 95-100% type 1 fibers, and in one of the 80% type I patients, some aspects of the muscle metabolism were studied by means of venous blood parameter analysis. Blood samples were derived from an antecubital vein prior to and at 0, 3, 5, and 7 minutes after the quadriceps test. Venous samples could not be taken directly from the m. quadriceps femoris. In these blood samples, the sodium, potassium, ammonium, and lactic acid levels are determined. Among these, lactic acid was the parameter most affected during ischemic exercise. The intramuscular lactate accumulation was thought to be an important cause of the decrease of the muscle membrane excitability during fatigue.".2"9'2 In controls, no blood samples were obtained during this ischemic quadriceps test. Therefore, reference values are not available. Additionally, the percentage of lactic acid rise was studied during the standardized ischemic forearm exercise test (IFET)"' in the same 5 patients. In this test, "arterialized" venous blood samples were obtained as can be deduced from the oxygen pressure (PO') rise in sequential venous blood samples during nonoccluded recovery. In controls, the percentage of lactate rise during IFET minimally exceeds a 300% increase in venous return blood."' EMG bursts were localized by determining their maxima after rectification and rectangularly windowing the EMG over 0.5 seconds. The power density frequency spectra of the subsequent EMG bursts were calculated by using the Fast Fourier Transform algorithm. T h e PDS was a representation of the distribution of frequencies that contributed to the EMG burst. Three parameters were derived: (1) from the power density spectrum the median frequency (Fmed) was determined. (2) The root mean square (RMS) value of the burst center was used as a measure of the SEMG amplitude. (3) The MFCV was determined by using a weighted linear regression on the phase differences between the two bi-
Data Analysis.
832
Fatigue in Type I Predominance
polar SEMG signals in series, considering the interelectrode distance of 12 mm.*O The change o f the parameters, as a function of time during the test, was expressed by the 2 linear regression parameters (intercept, slope). To account for the SEMG amplitude changes related to differences in the exerted force for each EMG burst separately, the "normalized amplitude" (amplitude/force [pV/N]) was calculated per EMG burst. By defining this operational parameter "normalized amplitude," we intended to study the force-generating capacity independently of the actually generated forces. The SEMG amplitude function is well characterized by a linear regression analysis. Spearman correlations and Student's t test were used for statistical evaluations. Results are given as mean values with (L) standard deviations in 95- 100% type I patients and in controls. In both 80% type I patients, individual data are presented. T h e overall individual data are given in Figure 2. RESULTS
(Table 1) The mean area of single type I fibers was increased compared with reference values in 6 of 8 patients at the time of biopsy. The mean area containing cores or nemaline rods ranged from 3% to 43%, with respect to the mean muscle fiber area. T h e percentage area of fat cells is clearly increased (1 1% to 5 1%) in all 95-100% type 1 fiber patients. In the patients with 80% type I fibers the percentage fat was, at most, 1%. See Table 1 for data on the number of fibers containing cores or rods. After combining these data concerning the reduction of contractile elements per cross-sectional area, the mean expected reduction of force ranged from 16% to 62% in patients with 95-100% type I fibers, and 0% to 7% in 80% type I patients.
Morphometric Studies.
Our data on the subcutaneous fat layer by means of skinfold measurements showed no significant differences between patients and controls (mean percentage fat of controls: 24.5 t 5.3%; 95-100% type I: 25.0 8.0%; and 80% type I: 20.7, 20.3, ns). T h e mean estimated upper arm muscle-bone area (MBA) in the 5 female patients with 95- 100% type I fibers (28.9 +- 4.7 cm2) was significantly lowered compared with female controls (45.3 4 3.2 cm2, P < 0.001). The mean MBA in male controls was 54.3 +- 9.9 cm'. On behalf of these mean MBA differences, the expected force
Skinfold and Upper Arm MBA Measurements.
*
MUSCLE & NERVE
September 1991
of patients with 95-100% type I fibers was further reduced to 65% of the force of controls. Force of M. Quadriceps Femoris. (Figure 2) The mean exerted force per kilogram of body weight in 95-100% type I predominance (0.85 ? 0.78 N/kg) was significantly less compared with 80% type I fibers (7.14 and 7.39 N/kg, P < 0.001) and compared with controls (7.36 ? 1.30 N/kg, P < 0.001). This means that 95- 100% type I patients showed a force reduction of 88% compared with controls. There were no significant differences concerning the force generation per kilogram of body weight of 80% type I patients and controls.
(Figure 2) Initially, the normalized SEMG amplitude in 95-100% type I predominance (2.64 k 1.35 FV/N) was significantly higher compared with controls (0.45 0.20 pV/N P < O.Ol), indicating a large difference in the force generation per microvolt of EMG amplitude between 95- 100% type I patients and controls. In 80% type I fibers, the initial normalized amplitudes were higher (0.76; 1.01 FV/N) than in controls (P < 0.05) and less than in patients with 95100% type I fibers, although significance limits were not reached. During the test, the mean slope of the normalized amplitude increased in controls (0.010 +- 0.005 s-'). In 95- 100% type I fibers, the normalized amplitude increased less (0.003 ? 0.004 s-', P < 0.01). In 80% type I fibers, one value representative for controls and one for 95100% type I fibers was found (0.01 1; 0.0 s-'). EMG Amplitude.
*
Patients with 95- 100% type I fibers showed a mean PT of 76 % 31 seconds during quadriceps exercise. Both 80% type I fiber patients performed 55 seconds. The mean PT in controls was 47 rt 8 seconds. Significance limits were not reached. The ischemic forearm exercise test (IFET) was performed by 4 95100% type I patients and by the male 80% type I patient. During IFET, the plasma concentrations of sodium, potassium, and ammonium showed no important changes different from reference values and need no further discussion. Patients with 95- 100% type I fibers showed lactate rises (32296, 71%, 358%, and 210% increases, respectively) during IFET, which are within the lower range of normal. During the ischemic qudriceps test, the percentage lactate rise in these patier its was very low or absent (O%, 5%, 0%, and 52% increases, respectively). However, the patient with 80% type I fibers had a normal lactate rise during IFET
Performance Time and Metabolism.
Fatigue in Type I Predominance
(550% increase). During the quadriceps test, the percentage lactate rise in this patient (120% rise) was high compared with the 95-100% type I patients. Muscle Fiber Conduction Velocity. (Figure 2) Patients with 80% type I fibers showed no significant differences concerning the initial MFCV (4.1; 4.2 m/s), neither to 95-100% type I patients (4.2 i. 0.7 m/s), nor to controls (5.5 ? 1.4 m/s). However, the mean initial MFCV of 95-100% type I fibers was significantly lower compared with controls ( P < 0.05). T h e MFCV of 95-100% type I fibers showed no change (mean slope: 0.0 0.005 m/s') during the test, which was significantly different from the mean slope of MFCV declination in controls (-0.013 k 0.012 m/s*, P < 0.05). The slope in MFCV declination in 80% type I patients (-0.007; -0.012 m/s') was lower compared with controls, and steeper compared with 95- 100% type I patients, although significance limits were not reached.
*
The mean initial Fmed of the PDS distribution did not differ significantly between 95- loo%, 80% type I fibers, and controls, although the mean initial Fmed of 95-100% type I patients (69.0 5 14.2 Hz) was somewhat lower compared with controls (77.7 2 13.5 Hz, ns). The mean slope of the Fmed decrease in 95-100% type I fibers (-0.17 & 0.21 Hz/s) was less pronounced, but not significantly different when compared with controls (-0.30 2 0.15 Hds, ns). T h e Fmed decrease in 80% type I fibers was -0.25 and -0.28 Hz/s. Power Density Frequency Spectrum (Figure 2).
DISCUSSION
(Table 1, Figure 2) As far as is known, the type I muscle fibers in congenital myopathies are normal type I fibers. Abnormalities are primarily sited at the core and rod zones, but the contractile structures and muscle membrane charateristics show no abnormalitie~.~~' Our data on the percentage fat, and on the area occupied by the cores and rods suggest that these changes may account for a mean reduction to approximately 63% (range 38% to 84%) of the muscle contractile elements of controls. The force actually generated by 95- 100% pathological type I fibers was approximately 12% of the force of controls. T h e data on the MBA showed that the upper arm muscle volume in patients with 95- 100% type I fibers was considerably reduced (65% of controls). In muscles with 80% type I fiForce.
MUSCLE & NERVE
September 1991
833
Force/Weig h t (N/W
interc Ampl/F (PV/N)
12 -
slope Amp1 (1/4 0.020 4
54 0
10 -
0
** B
0
-6-8-
I
z
4
‘I;
I
8 0
0
%
1 0 0
8 C 0 %
%
+
o.o*o;-/
~
------
--t -0.005
C
1 0
0 %
1 0 0
%
%
slope MFCV h/s21
interc Fmed (Hd
%
interc MFCV (m/s>
*
0
0
1 0
1
1 0
8 0
0
%
;
C
%
8 0
C
1 0
8 0
0
%
%
C
8 C 0 %
slope Fmed (Hz/s)
1
8
C
0 0 0 %
%
FIGURE 2. Scatter diagrams giving the force per kilogram of weight for the quadriceps femoris muscle, the intercepts (initial values), and slopes of the “normalized” amplitude (pV/N), the muscle fiber conduction velocity (MFCV), and the median frequency (Fmed) of the power density spectrum. Patients with 95-100% type l fibers (100%) are compared with patients with 80% type l fibers (80%), and to controls (C). 0 : male, *: female.
bers, the morphometric estimated reduction of contractile elements was less than 5%, and the MBA measurements in these patients showed no reduction of muscle volume. If w e assume that the core and rod zones do not generate force and that
834
Fatigue in Type I Predominance
the type I fibers basically act as normal, then the reduced force generating capacity in 95-100% type I patients originated from differences of the quadriceps muscle volume (ie, MBA) and/or from the absence of type I1 fibers.
MUSCLE & NERVE
September 1991
We are aware that these results must be interpreted with care. Not all the muscle biopsies are taken from the quadriceps femoris muscle; the exact area of rods is hard to calculate in fibers where rods are diffusely distributed; and there is a varying time span between the force measurements and biopsy date. Furthermore, biopsy specimens taken for diagnosis must, preferably, contain not too much fat, which will result in an underestimation of the percentage of fat. Considering the assumptions made, our data on the expected reduction of contractile elements (63%), the reduced muscle volume (65%), an underestimated fat content (eg, 90% of controls), and the fact that normal quadriceps muscle contains about 50% type I fibers will lead to the conclusion that type I1 muscle fibers are approximately 5 times stronger than type I fibers. This result supports the findings of Young et aLS5 who reported that human quadriceps type I1 fibers are 3 times stronger than type I fibers. (Figure 2) The EMG amplitude is often considered to be a reflection of the central drive4 and, normally, the EMG amplitude increases nearby proportionally with the exerted for~e.~ In, our ~ , ~experiment, ~ the initial normalized SEMG amplitude [kV/N] was about 6 times higher in patients with 95-100% type I fibers compared with controls. When more EMG voltage is needed per unit of force, this means that what can be called the neuromuscular efficiency “ME] of type I fibers is worse. This finding is consistent with the above prediction on the lower forcegenerating capacity of type I muscle fibers. Another finding is that, during the test, the normalized amplitude increased less in 95- 100% type I fibers compared with controls and in 1 of the patients with 80% type I fibers, while the force was kept constant. This finding stresses the relative resistance to fatigue of type I fibers, although it may be that 95- 100% type I fiber patients need nearly full motor unit recruitment to obtain SO% of MVC. In that case, one would expect only small changes in the normalized amplitude. The increase of the SEMG amplitude in controls is thought to result from some additional recruitment of type I1 fibers and an increasing motor unit firing rate.
EMG Amplitude.
The metabolism of type I fibers is mainly aerobic,“ and these fibers show a lower energy cost for calcium cross-bridge cycling than type I I fibers,”
Performance Time and Lactate Formation.
Fatigue in Type I Predominance
which is what makes the type I muscle fibers more resistant to fatigue. T h e PT results indicated that, under anerobic conditions, this fatigue resistance is reduced to a level comparable with that of type I1 fibers. During IFET the lactic acid production was low in 95-100% type I patients, and normal in the 20% type I1 fibers patient. During the quadriceps test, only the 20% type I1 patient offered a substantial contribution of lactate to the general circulation. This confirms the general conclusion that the lactate formation during anerobic exercise is a function of the recruitment of type 11 fibers.6,8.15,22,26
Muscle Fiber Conduction Velocity. (Figure 2) In contrast with the findings concerning the force level and PT, where the presence of type I1 fibers determines the results, we found the initial MFCV to be a parameter in which type I fiber predominance was the determining factor. The mean MFCV in 95- 100% type I fiber patients (4.2 m/s) and the MFCV in patients with 80% type I fibers were almost equal (4.1; 4.2 m/s). We would have expected that the MFCV of 80% type I fibers to be intermediate between controls and 95- 100% type I fibers. Usually, the MFCV is thought to be related to the diameter of the muscle but in animal studies18 no differences concerning the MFCV of type I and I1 fibers are found. Sadoyama et al.*’ showed that the MFCV is related to the area of the type I1 fibers per cross-sectional area, irrespective of the muscle fiber diameter. In most of our 95-100% type I patients, the mean fiber area of the type I fibers was considerably increased, and even larger than that of normal type I1 fibers.35 This showed that the differences concerning the initial MFCV are not caused by a decreased mean fiber diameter. This result from hypertrophic type I fibers in congenital myopathies gives full support to the conclusions of Sadoyama et a P 9 A most significant finding was that the conduction velocity along the membrane of 95- 100% type I fibers showed no declination during exercise. With an increasing percentage of type I1 fibers, the mean negative slope of MFCV increased. Because the MFCV is a sensitive neurophysiological sign of the muscle membrane e x ~ i t a b i l i t y , ~ ~ ’ ~ ~ ~ ~ this indicates that during anerobic exercise the excitability of muscle membrane in type I1 fibers becomes more reduced than in type I fibers. Generally, the intramuscular accumulation of lactic
MUSCLE & NERVE
September 1991
835
acid is held responsible for the slowing of MFCV,6,8,z0,2zwhich is in accordance with the findings from IFET during which patients who possess type I1 fibers, indeed, produce more lactic acid. It remains uncertain whether the MFCV of type I fibers will decrease in normal muscle where these fibers are surrounded by lactic acidproducing type I1 fibers. (Figure 2) The initial median frequency of the PDS of 95- 100% type I fibers was minimally lowered, and the shift of PDS during fatigue was less compared with controls. None of these findings reached significant limits. T h e somewhat-lowered mean initial Fmed cannot be explained by the volumeconducting characteristics of the subcutaneous fat layer, acting as a low-pass because the mean percentage fat distribution did not differ significantly between patients and controls. Our findings support the earlier observations that muscles with a higher percentage of type I1 fibers generate a higher Fmedz4 and that the Fmed depends on the MFCV.','* Many authors reported on a shift of PDS to lower frequency in muscle fatigue.3,8,12,1636 I n our experiment, we found a definite shift of Fmed to lower frequency in both patients and controls. This shift of PDS was slightly related to the proportion of fiber types. Two major mechanisms are thought to cause this shift: decrease of the MFCV and changing of the motor unit firing statistics, eg, synchronization.8,z3725T h e slightly lowered shift of PDS in 95- 100% type I fibers can be explained by the absence of MFCV declination in these fibers. Some authors reported that the percentage fall
Power Density Frequency Spectrum.
of Fmed always exceeds that of MFCV,6*36while others found a pro ortional relationship between both parameters. 8,1g19,22 Our findings were in accordance with the first group of findings, and suggest once morezo that the MFCV cannot be the only factor determining the shift of PDS, and that the motor unit firing statistics cause the PDS to shift to lower frequencies. We found no evidence that the influence of the statistics was stronger in patients with 95- 100% type I fibers since, in these patients, the mean value of the negative slope of Fmed can be explained mathematically by the absence of MFCV declination when compared with controls. In conclusion, in patients with type I fiber predominance it was shown that type I muscle fibers have a lower force-generating capacity and a lower MFCV compared with controls. The absence of a MFCV declination in type I fibers indicates an unimpaired membrane excitability during ischemic exercise at 80% MVC, which appears related to the lack of a substantial lactic acid formation by type I fibers. Additional support for our findings originates from the results in patients with 20% type I1 fibers who, in general, show intermediate values between controls and 95- 100% type I fibers. Type I muscle fibers fatigue despite the lack of definite PDS and MFCV changes in the surface electromyogram. T h e EMG amplitude findings also indicate no electrophysiological fatigability. Therefore, these findings support the conclusion that muscle fatigue in patients with 95- 100% type I fiber predominance originates beyond the scope of extracellular SEMG measurements at a sarcolemmal level.
REFERENCES 1. Arendt-Nielsen L, Forster A, Mills KR: The relationship between niuscle fiber conduction velocity and force in human vastus lateralis. J Physiol 1984;353:6P. 2. Banker BQ: The congenital myopathies, in Engel AG, Banker BQ (eds): Myology. New York, McGraw-Hill, 1986, pp 1527-1581. 3. Bigland-Ritchie €3, Donovan EF, Roussos CS: Conduction velocity and EMG power density changes in fatigue of sustained maximal efforts. J Appl Physiol 198 1;51: 1300- 1305. 4. Bigland-Ritchie B, Woods JJ: Changes in muscle contractile properties and neural control during human muscular fatigue. Muscle Nerve 1984;7:691-699. 5 . Braakhekke J P , Stegeman DF, Joosten EMG: Increase in median power frequency of the myoelectric signal in pathological fatigue. Electroenceph Clan Neurophysiol 1989;73:151- 156. 6. Broman H, Bilotto G, DeLuca C: Myoelectric signal conduction velocity and spectral parameters: influence of force and time. J Appl Physiol 1985;58: 1428- 1437.
836
Fatigue in Type I Predominance
7. Brooke MH: Congenital (more or less) muscle diseases, in Brooke MH (ed): A Clinician's Vzew of Neuromuscular DiSeuses. Baltimore, Williams & Wilkins, 1986, pp 340-380. 8. DeLuca CJ: Myoelectrical manifestations of localised muscular fatigue in humans. CRC Crit Rev Bioeng 1984; 112 5 1-279. 9. Donselaar Y, Eerbeek 0, Kernell D, Verhey BA: Fibre sizes and histochemical staining characteristics in normal and chronically stimulated fast muscle in cat. J Appl Physiol 1987;382:237-254. 10. Dubowitz V, Brooke MH: Definition of pathological changes, in Dubowitz V and Brooke MH (eds): Muscle biopsy: A Modern Approach. London-Philadelphia-Toronto, Saunders, 1973, pp 74- 102. 11. Durnin JVGA, Womersley J: Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 7 2 years. 5r J Nutr 1974;32:77-97. 12. Eberstein A, Beattie B: Simultaneous measurement of con-
MUSCLE & NERVE
September 1991
duction velocity and EMG power spectrum changes during fatigue. Muscle Nerve 1985;8:768-773. 13. Hatcher DD, Luff AR: Force-velocity properties of fasttwitch and slow-twitch muscles of the' k;tten. J Physiol 1985;367:377-385. 14. Horita T , Ishiko T: Relationships between muscle lactate accumulation and surface EMG activities during isokinetic contractions in man. Eur J Appl Physiol 1987;56: 18-23. 15. Juel C: Muscle action potential propagation velocity changes during activity. Muscle Nerve 1988;11:714-719. 16. Kereshi S, Manzano G , McComas AJ: Impulse conduction velocities in human biceps branchii muscles. Exp Neural 1983;80:652- 662. 17. Koning FL de, Rinkhorst RA, Kauer JMG, Thijssen HOM: Accuracy of an anthropometric estimate of the muscle and bone area in a transversal cross-section of the arm. Int J Sports Med 1986;7:246-249. 18. K.sler F, Lange F, Caffier G, Kuchler G: Changes of the conduction velocity of isolated muscles induced by altered external potassium concentration. Biomed Biochim Acta 1989;48:465- 470. 19. Lindstrdm L, Peterson I: Power spectrum of EMG signals and its applications, in Desmedt JE (ed): Progress in Clinicul Neurophysiology: Computer-aided Elrctromyopaphy, vol 10. Basel, Karger, 1983, pp 1-51. 20. Linssen WHJP, Jacobs M, Stegeman DF, et al: Fatigue in McArdle's disease. Surface EMG frequency spectrum and muscle fiber conduction velocity during ischemic exercise. Bruin 1990;113:1779- 1793. 21. Maughan RJ, Nimmo MA: T h e inHuence of variations in muscle fibre composition on muscle strength and cross-sectional area in untrained males. J Physiol 1984;351:29931 1. 22. Merletti R, Sabbahi MA, DeLuca CJ: Median frequency of the myoelectric signal. Effects of muscle ischemia and cooling. EurJ Appl Physiol 1984;52:258-265. 23. Miller RG, Giannini D, Milner-Brown HS, et al: Effects of fatiguing exercise on high energy phosphates, force, and EMG: evidence of three phases of recovery. Muscle Nerve 1987; 10:8 10-8'21.
Fatigue in Type I Predominance
24. Milner-Brown HS, Mellenthin M, Miller RG: Quantifying human muscle strength, endurance and fatigue. Arch P I y Mrd Rehabil 1986;67:530-535. 25. Mills KR, Edwards RHT: Muscle fatigue in myophosphorylase deficiency; power spectral analysis of the electromyogram. Electroenceph Clin Neurophysiol 1984;57:330-335. 26. Moritani T, Tanaka H, Yoshida T, et al: Relationship between myoelectric signals and blood lactate during incremental forearm exercise. Am,]Phys Med 1984;63: 122- 132. 27. Ruegg JC: Excitation-contraction coupling in fast- and slow-twitch muscle fibers. Int J Sports Med 1987;8:360-364. 28. Ruff RL: Calcium sensitivity of fast- and slow-twitch human muscle fibers. Muscle Nerve 1989;12:32-37. 29. Sadoyama T, Masuda T, Miyata H, Katsuda S: Fibre cow duction velocity and fibre composition in human vastus lateralis. Eur J Appl Physiol 1988;57:767-771. 30. Sale DG: Influence of exercise and training on motor unit activation. Exerc Sporl Sci Rev 1987;15:95- 151. 31. Schantz P, Randall-Fox E, Hutchison W, Tyden A, Astrand PO: Muscle fibre type distribution, muscle cross-sectional area and maximal voluntary strength in humans. Acta Physiol Scand 1983;117:219-226. 32. Sinkeler SPT, Wevers RA, Joosten EMG, et al: Improvement of screening in exertional myalgia with a standardized ischemic forearm test. Muscle Neiur 1986;9:731-737. 33. Sinkeler SPT: Myoadenylaat deaminase deficiency. Thesis, Nijmegen, T h e Netherlands, 1987, pp 59-70. 34. Yoneda T, Oishi K, Fujikura S, Ishida A: Recruitment threshold force and its changing type of motor units during voluntary contraction at various speeds in man. Bruin Res 1986;372:89-94. 35. Young A: T h e relative isometric strength of type I and type 2 muscle fibres in the human quadriceps. Clin Physiol 1984;4:23-32. 36. Zwarts MJ, Van Weerden TW, Haenen HTM: Relation between average muscle fibre conduction velocity and EMG power spectra during isometric contraction, recovery and applied ischemia. Eur J Appl Physiol 1987;59:212-216.
MUSCLE & NERVE
September 1991
837
FATIGUE IN TYPE I FIBER PREDOMINANCE: A MUSCLE FORCE AND SURFACE EMG STUDY ON THE RELATIVE ROLE OF TYPE I AND TYPE II MUSCLE FIBERS WIM H.J.P. LINSSEN, MD, DICK F. STEGEMAN, PhD, Ed M.G. JOOSTEN, MD, PhD, ROB A. BINKHORST, PhD, MIEKE J.H. MERKS, BSc, HENK J. TER LAAK, PhD, and SERVAAS L.H. NOTERMANS, MD, PhD
T h e congenital myopathies (ie, central core disease [CCD] and nemaline rod myopathy [NRM]) are characterized by a generalized type I fiber predominance in all skeletal muscles, except for some muscles who are innervated by cranial nerves2,' The relative proportions of fiber types within From the Institutes of Neurology (Drs. Linssen, Stegeman, Joosten, Merks, ter Laak, and Notermans) and Physiology (Dr. Binkhorst). University of Nijmegen, The Netherlands. This investigation is part of the research program "Disorders of the neurornuscular system" of the University of Nijmegen. Acknowledgments: The authors thank Mrs. G. Steenbergen and Mrs. L. van Woerkom for their technical assistance and the laboratory examinations, and Mr. H. Wijnen and Mr. N. Dijkstra for preparing the figures. Address reprint requests to W.H.J.P. Linssen, MD, Institute of Neurology, University of Nijrnegen. P.O. Box 9101, 6500 HB Nijmegen, The Netherlands. Accepted for publication August 14, 1990.
CCC 0148-639X/91/090829-09 $04.00 0 1991 John Wiley & Sons, Inc.
Fatigue in Type I Predominance
muscle and the characteristics of these muscle fiber types are undoubtedly important determinants of the surface electromyogram (SEMG) during fatiguing exercise. Normally type I muscle fibers become recruited at low force levels. When the force level is increased, type I1 motor units become recruited in addition.34 The abovementioned congenital myopathies may serve as a model to study the type I fiber characteristics. Type I1 motor units show a fast action potential conduction velocity along the muscle fiber membrane (MFCV). This explains why the MFCV increases with increasing force levels.6229The larger MFCV is thought to be related to the larger diameter of type I1 fibers,6,8,'6although Sadoyama et al?' found the MFCV not positively correlated with the muscle fiber diameter. In their experiment, the MFCV is related to the cross-sectional area of type I1 fibers. During fatiguing exercise, the power density
MUSCLE & NERVE
September 1991
829
frequency spectrum (PDS) shifts to lower frequenc i e ~ . ' ~Milner-Brown '~ et al.24 showed that (unfatigued) type I1 fibers have a higher median frequency. Braakhekke et al.5 found the PDS shifting toward higher frequencies in a patient suffering from a carnitine deficiency during submaximal (30% of the maximal aerobic power) bicycle ergometric exercise over a 2-hour period. From these experiments it was suggested that, like the MFCV, the PDS and its shift during fatigue were related to the relative role of type I1 and type I fibers. Another aspect is whether type I muscle fibers generate more force than type I1 fibers. Some authors find no significant differences between the force-generating capacity of human type I and type I1 muscle fibers"; others find type I1 muscle fibers more powerful.35 Some studies2123'suggest that strength is correlated to the muscle crosssectional area irrespective of the fiber composition. Animal s t ~ d i e s ~ ~ have ' " ' ~ shown that type I1 fibers are capable of generating more force than type I fibers. T h e present study had two aims. By studying the above-described congenital myopathies, it became possible: (a) to determine the forcegenerating capacity, the MFCV and SEMG frequency spectrum parameters of muscle depending on the fiber type composition; and (b) to investigate the changes of the MFCV, the SEMG frequency spectrum and amplitude during anerobic fatiguing exercise, again depending on the fiber type composition and some metabolic parameters. MATERIALS AND METHODS Patients and Controls. six mean age 23.4 ? 7.6 years,
patients (5 females: mean height 1.58 0.08 m, mean weight 46.8 5 8.7 kg; 1 male: 13 years; height 1.40 m, weight 29 kg) with 95- 100% type I fiber predominance together with two patients (1 female: 15 years, 1.64 m, 40 kg; l male: 35 years, 1.74 m, 72 kg) with SO% type I muscle fibers in their muscle biopsy specimen performed isometric ischemic intermittent m. quadriceps femoris contractions at 80% maximal voluntary contraction (MVC), rate 30/min. Twelve healthy volunteers (8 females: mean age 30 k 6.4 years, mean height 1.71 2 0.03 m, mean weight 60 ? 6.8 kg; 4 males: mean age 36 ? 11.3 years, mean height 1.79 ? 0.16, mean weight 74.5 ? 10.2 kg) served as a control group. As all controls were without neuromuscular complaints, no muscle biopsies were taken. A normal mixed distribution with more than 45% type I1 muscle fibers within the m. quadriceps femoris was assumed." All pa-
830
Fatigue in Type I Predominance
*
tients and control subjects gave their informed consent. The Committee on Experiments in Humans of the Faculty of Medicine saw and approved the experimental protocol. (Table 1) Of the patients with 95100% type I fibers, 5 were suffered from central core disease (CCD), and 1 patient from adult autosoma1 dominant nemaline rod myopathy (NRM). Of the 2 patients with 80% type I fibers, one suffered from CCD. The other patient complained of exertional myalgia symptoms. In her biopsy specimen, no structural abnormalities were found except for a marked type I fiber predominance. In the congenital myopathies, the amount of structural abnormalities (cores and rods) and fat cells may vary per muscle group.' In most cases, the muscle biopsies were taken from the m. quadriceps femoris. Morphometric studies were carried out using SDH or NADH-TR stained sections. Area measurements of the muscle fibers and of the unstained parts (representing the areas containing cores or nemaline rods) of these fibers were performed by using a systematic random point sampling design. In this way, the mean fiber area, the relative area of cores or rods, the percentage of fibers containing cores, and the percentage of area of fat cells were calculated. The area measurements were adjusted to age at the time of biopsy. T o relate the physiological results to the gross anatomical structures of the limb, skinfold measurements (percentage fat)' l were carried out, and the mean anthropometric estimated upper arm muscle-bone area (MBA [cm2])" was determined in most subjects. T h e MBA measurements were standardized for the upper arm muscles and, therefore, offered only indirect information about the muscle volume of the m. quadriceps femoris. Biopsy Data.
Experimental Set-Up. T h e experimental protocol of the isometric ischemic m. quadriceps femoris test resembled the ischemic forearm exercise test (IFET) as described by Sinkeler et al.32 In short, all subjects were seated upright on a table with the upper leg supported and the lower leg hanging freely. The ankle was firmly attached to a strain gauge dynamometer (Fig. lA).3" The maximal voluntary contraction level (MVC) of the right m. quadriceps femoris was determined twice with 1 minute of rest in between. T h e amplified force signal was recorded on paper. Thereafter, all subjects relaxed for 10 minutes. During the test, intermittent contractions were
MUSCLE & NERVE
September 1991
~~~
~
~
~
~
Table 1. Morphometric studies 80% Type I
95-100% Type I
Sexlage (yr) Biopsy site Type lldisease (%)
Number of fibers counted Mean fiber* area (pm2) Mean area with coreslrods per fiber (%) Area with fat cells (%) Estimated reduction of contractile elements (%)
1
2
3
4
5
6
7
8
MI13* Erec. tr 100 CCD
Fl15 quadr 100 CCD
F/21t quadr. 96 CCD
Fl20
soleus 100 CCD
F126 quadr 100 CCD
Fl35 quadr 100 NRM
MI35 quadr 80 CCD
F/l5 quadr. 80 type 1
112 809 9
I34 4321 14 11 23
72 936 3 14 16
112 2208 21 11 30
7 3226 43 33 62
68 4601 29 13 38
126 8052 7
75 5398 0 0.5 0
51
55
1 7
‘Biopsy at 7 year of age. f8iopsy at 5 years of age. #Calculations are made by statisticaliy transformation of the muscle fiber dmmeter variabie ( d ) into a new variable (a), so that a = ~rd’/4.In adult controls we suppose” that d has a normal distribution of about 60 pm’ and a variance of 64 pm’. It follows that the calculated normal mean fiber area (a) will be 2878 t 757 pm’. In our laboratory, the mean fiber diameter in normal subjects is assumed to be 20 ? 3 pm at age I and 30 2 5 pm at age 5 . It follows a normal mean fiber area of 327 2 95 pm’ at age 1, and 726 2 237 pm’ at age 5 .
made at 80% of MVC under ischemic conditions. Ischemia was applied by inflation of a cuff above 200 mmHg pressure 5 seconds prior to the experiment. The cuff was wrapped around the m. quadriceps as proximal as possible. Intermittent contractions were made at a rate of 30 per minute until the 80% MVC force level could no longer be reached. This moment was defined as the failure point. Data analysis was performed over the performance time (PT)ending at the failure point.
A
Fatigue in Type I Predominance
Much care was taken in positioning the SEMG array parallel to the direction of the muscle fibers of the short head of right m. vastus medialis, distal to the motor point (Fig. 1A). We used a specially designed surface electrode array consisting of 4 gold-coated electrodes placed in-line. The electrode diameter was 2 mm and the interelectrode distance was 6 mm (Fig. IB). The skin was carefully abrased and cleaned with ethanol. N o electrode paste was used in order to avoid low-
FIGURE 1. (A) Subjects are seated and firmly attached to the quadriceps dynamometer. The right leg is horizontally fixed. The lower leg is hanging freely. The surface electrode array is positioned distally from the motor point on the short head of the m. vastus medialis. (B) Detail of the surface electrode array consisting of 5 electrodes. Four electrodes are used for deriving 2 bipolar SEMG signals.
MUSCLE & NERVE
September 1991
831
resistance bridges between the closely spaced electrodes. The experimental protocol, the way the SEMG signals were derived, and the data analysis have been described previously."' After amplification, band pass filtering, AID conversion, sampling, and storing on magnetic disk, the SEMG signals were compared by determining the cross-correlation function. The maximum cross-correlation should exceed 0.7. In 4 patients with 95-100% type 1 fibers, and in one of the 80% type I patients, some aspects of the muscle metabolism were studied by means of venous blood parameter analysis. Blood samples were derived from an antecubital vein prior to and at 0, 3, 5, and 7 minutes after the quadriceps test. Venous samples could not be taken directly from the m. quadriceps femoris. In these blood samples, the sodium, potassium, ammonium, and lactic acid levels are determined. Among these, lactic acid was the parameter most affected during ischemic exercise. The intramuscular lactate accumulation was thought to be an important cause of the decrease of the muscle membrane excitability during fatigue.".2"9'2 In controls, no blood samples were obtained during this ischemic quadriceps test. Therefore, reference values are not available. Additionally, the percentage of lactic acid rise was studied during the standardized ischemic forearm exercise test (IFET)"' in the same 5 patients. In this test, "arterialized" venous blood samples were obtained as can be deduced from the oxygen pressure (PO') rise in sequential venous blood samples during nonoccluded recovery. In controls, the percentage of lactate rise during IFET minimally exceeds a 300% increase in venous return blood."' EMG bursts were localized by determining their maxima after rectification and rectangularly windowing the EMG over 0.5 seconds. The power density frequency spectra of the subsequent EMG bursts were calculated by using the Fast Fourier Transform algorithm. T h e PDS was a representation of the distribution of frequencies that contributed to the EMG burst. Three parameters were derived: (1) from the power density spectrum the median frequency (Fmed) was determined. (2) The root mean square (RMS) value of the burst center was used as a measure of the SEMG amplitude. (3) The MFCV was determined by using a weighted linear regression on the phase differences between the two bi-
Data Analysis.
832
Fatigue in Type I Predominance
polar SEMG signals in series, considering the interelectrode distance of 12 mm.*O The change o f the parameters, as a function of time during the test, was expressed by the 2 linear regression parameters (intercept, slope). To account for the SEMG amplitude changes related to differences in the exerted force for each EMG burst separately, the "normalized amplitude" (amplitude/force [pV/N]) was calculated per EMG burst. By defining this operational parameter "normalized amplitude," we intended to study the force-generating capacity independently of the actually generated forces. The SEMG amplitude function is well characterized by a linear regression analysis. Spearman correlations and Student's t test were used for statistical evaluations. Results are given as mean values with (L) standard deviations in 95- 100% type I patients and in controls. In both 80% type I patients, individual data are presented. T h e overall individual data are given in Figure 2. RESULTS
(Table 1) The mean area of single type I fibers was increased compared with reference values in 6 of 8 patients at the time of biopsy. The mean area containing cores or nemaline rods ranged from 3% to 43%, with respect to the mean muscle fiber area. T h e percentage area of fat cells is clearly increased (1 1% to 5 1%) in all 95-100% type 1 fiber patients. In the patients with 80% type I fibers the percentage fat was, at most, 1%. See Table 1 for data on the number of fibers containing cores or rods. After combining these data concerning the reduction of contractile elements per cross-sectional area, the mean expected reduction of force ranged from 16% to 62% in patients with 95-100% type I fibers, and 0% to 7% in 80% type I patients.
Morphometric Studies.
Our data on the subcutaneous fat layer by means of skinfold measurements showed no significant differences between patients and controls (mean percentage fat of controls: 24.5 t 5.3%; 95-100% type I: 25.0 8.0%; and 80% type I: 20.7, 20.3, ns). T h e mean estimated upper arm muscle-bone area (MBA) in the 5 female patients with 95- 100% type I fibers (28.9 +- 4.7 cm2) was significantly lowered compared with female controls (45.3 4 3.2 cm2, P < 0.001). The mean MBA in male controls was 54.3 +- 9.9 cm'. On behalf of these mean MBA differences, the expected force
Skinfold and Upper Arm MBA Measurements.
*
MUSCLE & NERVE
September 1991
of patients with 95-100% type I fibers was further reduced to 65% of the force of controls. Force of M. Quadriceps Femoris. (Figure 2) The mean exerted force per kilogram of body weight in 95-100% type I predominance (0.85 ? 0.78 N/kg) was significantly less compared with 80% type I fibers (7.14 and 7.39 N/kg, P < 0.001) and compared with controls (7.36 ? 1.30 N/kg, P < 0.001). This means that 95- 100% type I patients showed a force reduction of 88% compared with controls. There were no significant differences concerning the force generation per kilogram of body weight of 80% type I patients and controls.
(Figure 2) Initially, the normalized SEMG amplitude in 95-100% type I predominance (2.64 k 1.35 FV/N) was significantly higher compared with controls (0.45 0.20 pV/N P < O.Ol), indicating a large difference in the force generation per microvolt of EMG amplitude between 95- 100% type I patients and controls. In 80% type I fibers, the initial normalized amplitudes were higher (0.76; 1.01 FV/N) than in controls (P < 0.05) and less than in patients with 95100% type I fibers, although significance limits were not reached. During the test, the mean slope of the normalized amplitude increased in controls (0.010 +- 0.005 s-'). In 95- 100% type I fibers, the normalized amplitude increased less (0.003 ? 0.004 s-', P < 0.01). In 80% type I fibers, one value representative for controls and one for 95100% type I fibers was found (0.01 1; 0.0 s-'). EMG Amplitude.
*
Patients with 95- 100% type I fibers showed a mean PT of 76 % 31 seconds during quadriceps exercise. Both 80% type I fiber patients performed 55 seconds. The mean PT in controls was 47 rt 8 seconds. Significance limits were not reached. The ischemic forearm exercise test (IFET) was performed by 4 95100% type I patients and by the male 80% type I patient. During IFET, the plasma concentrations of sodium, potassium, and ammonium showed no important changes different from reference values and need no further discussion. Patients with 95- 100% type I fibers showed lactate rises (32296, 71%, 358%, and 210% increases, respectively) during IFET, which are within the lower range of normal. During the ischemic qudriceps test, the percentage lactate rise in these patier its was very low or absent (O%, 5%, 0%, and 52% increases, respectively). However, the patient with 80% type I fibers had a normal lactate rise during IFET
Performance Time and Metabolism.
Fatigue in Type I Predominance
(550% increase). During the quadriceps test, the percentage lactate rise in this patient (120% rise) was high compared with the 95-100% type I patients. Muscle Fiber Conduction Velocity. (Figure 2) Patients with 80% type I fibers showed no significant differences concerning the initial MFCV (4.1; 4.2 m/s), neither to 95-100% type I patients (4.2 i. 0.7 m/s), nor to controls (5.5 ? 1.4 m/s). However, the mean initial MFCV of 95-100% type I fibers was significantly lower compared with controls ( P < 0.05). T h e MFCV of 95-100% type I fibers showed no change (mean slope: 0.0 0.005 m/s') during the test, which was significantly different from the mean slope of MFCV declination in controls (-0.013 k 0.012 m/s*, P < 0.05). The slope in MFCV declination in 80% type I patients (-0.007; -0.012 m/s') was lower compared with controls, and steeper compared with 95- 100% type I patients, although significance limits were not reached.
*
The mean initial Fmed of the PDS distribution did not differ significantly between 95- loo%, 80% type I fibers, and controls, although the mean initial Fmed of 95-100% type I patients (69.0 5 14.2 Hz) was somewhat lower compared with controls (77.7 2 13.5 Hz, ns). The mean slope of the Fmed decrease in 95-100% type I fibers (-0.17 & 0.21 Hz/s) was less pronounced, but not significantly different when compared with controls (-0.30 2 0.15 Hds, ns). T h e Fmed decrease in 80% type I fibers was -0.25 and -0.28 Hz/s. Power Density Frequency Spectrum (Figure 2).
DISCUSSION
(Table 1, Figure 2) As far as is known, the type I muscle fibers in congenital myopathies are normal type I fibers. Abnormalities are primarily sited at the core and rod zones, but the contractile structures and muscle membrane charateristics show no abnormalitie~.~~' Our data on the percentage fat, and on the area occupied by the cores and rods suggest that these changes may account for a mean reduction to approximately 63% (range 38% to 84%) of the muscle contractile elements of controls. The force actually generated by 95- 100% pathological type I fibers was approximately 12% of the force of controls. T h e data on the MBA showed that the upper arm muscle volume in patients with 95- 100% type I fibers was considerably reduced (65% of controls). In muscles with 80% type I fiForce.
MUSCLE & NERVE
September 1991
833
Force/Weig h t (N/W
interc Ampl/F (PV/N)
12 -
slope Amp1 (1/4 0.020 4
54 0
10 -
0
** B
0
-6-8-
I
z
4
‘I;
I
8 0
0
%
1 0 0
8 C 0 %
%
+
o.o*o;-/
~
------
--t -0.005
C
1 0
0 %
1 0 0
%
%
slope MFCV h/s21
interc Fmed (Hd
%
interc MFCV (m/s>
*
0
0
1 0
1
1 0
8 0
0
%
;
C
%
8 0
C
1 0
8 0
0
%
%
C
8 C 0 %
slope Fmed (Hz/s)
1
8
C
0 0 0 %
%
FIGURE 2. Scatter diagrams giving the force per kilogram of weight for the quadriceps femoris muscle, the intercepts (initial values), and slopes of the “normalized” amplitude (pV/N), the muscle fiber conduction velocity (MFCV), and the median frequency (Fmed) of the power density spectrum. Patients with 95-100% type l fibers (100%) are compared with patients with 80% type l fibers (80%), and to controls (C). 0 : male, *: female.
bers, the morphometric estimated reduction of contractile elements was less than 5%, and the MBA measurements in these patients showed no reduction of muscle volume. If w e assume that the core and rod zones do not generate force and that
834
Fatigue in Type I Predominance
the type I fibers basically act as normal, then the reduced force generating capacity in 95-100% type I patients originated from differences of the quadriceps muscle volume (ie, MBA) and/or from the absence of type I1 fibers.
MUSCLE & NERVE
September 1991
We are aware that these results must be interpreted with care. Not all the muscle biopsies are taken from the quadriceps femoris muscle; the exact area of rods is hard to calculate in fibers where rods are diffusely distributed; and there is a varying time span between the force measurements and biopsy date. Furthermore, biopsy specimens taken for diagnosis must, preferably, contain not too much fat, which will result in an underestimation of the percentage of fat. Considering the assumptions made, our data on the expected reduction of contractile elements (63%), the reduced muscle volume (65%), an underestimated fat content (eg, 90% of controls), and the fact that normal quadriceps muscle contains about 50% type I fibers will lead to the conclusion that type I1 muscle fibers are approximately 5 times stronger than type I fibers. This result supports the findings of Young et aLS5 who reported that human quadriceps type I1 fibers are 3 times stronger than type I fibers. (Figure 2) The EMG amplitude is often considered to be a reflection of the central drive4 and, normally, the EMG amplitude increases nearby proportionally with the exerted for~e.~ In, our ~ , ~experiment, ~ the initial normalized SEMG amplitude [kV/N] was about 6 times higher in patients with 95-100% type I fibers compared with controls. When more EMG voltage is needed per unit of force, this means that what can be called the neuromuscular efficiency “ME] of type I fibers is worse. This finding is consistent with the above prediction on the lower forcegenerating capacity of type I muscle fibers. Another finding is that, during the test, the normalized amplitude increased less in 95- 100% type I fibers compared with controls and in 1 of the patients with 80% type I fibers, while the force was kept constant. This finding stresses the relative resistance to fatigue of type I fibers, although it may be that 95- 100% type I fiber patients need nearly full motor unit recruitment to obtain SO% of MVC. In that case, one would expect only small changes in the normalized amplitude. The increase of the SEMG amplitude in controls is thought to result from some additional recruitment of type I1 fibers and an increasing motor unit firing rate.
EMG Amplitude.
The metabolism of type I fibers is mainly aerobic,“ and these fibers show a lower energy cost for calcium cross-bridge cycling than type I I fibers,”
Performance Time and Lactate Formation.
Fatigue in Type I Predominance
which is what makes the type I muscle fibers more resistant to fatigue. T h e PT results indicated that, under anerobic conditions, this fatigue resistance is reduced to a level comparable with that of type I1 fibers. During IFET the lactic acid production was low in 95-100% type I patients, and normal in the 20% type I1 fibers patient. During the quadriceps test, only the 20% type I1 patient offered a substantial contribution of lactate to the general circulation. This confirms the general conclusion that the lactate formation during anerobic exercise is a function of the recruitment of type 11 fibers.6,8.15,22,26
Muscle Fiber Conduction Velocity. (Figure 2) In contrast with the findings concerning the force level and PT, where the presence of type I1 fibers determines the results, we found the initial MFCV to be a parameter in which type I fiber predominance was the determining factor. The mean MFCV in 95- 100% type I fiber patients (4.2 m/s) and the MFCV in patients with 80% type I fibers were almost equal (4.1; 4.2 m/s). We would have expected that the MFCV of 80% type I fibers to be intermediate between controls and 95- 100% type I fibers. Usually, the MFCV is thought to be related to the diameter of the muscle but in animal studies18 no differences concerning the MFCV of type I and I1 fibers are found. Sadoyama et al.*’ showed that the MFCV is related to the area of the type I1 fibers per cross-sectional area, irrespective of the muscle fiber diameter. In most of our 95-100% type I patients, the mean fiber area of the type I fibers was considerably increased, and even larger than that of normal type I1 fibers.35 This showed that the differences concerning the initial MFCV are not caused by a decreased mean fiber diameter. This result from hypertrophic type I fibers in congenital myopathies gives full support to the conclusions of Sadoyama et a P 9 A most significant finding was that the conduction velocity along the membrane of 95- 100% type I fibers showed no declination during exercise. With an increasing percentage of type I1 fibers, the mean negative slope of MFCV increased. Because the MFCV is a sensitive neurophysiological sign of the muscle membrane e x ~ i t a b i l i t y , ~ ~ ’ ~ ~ ~ ~ this indicates that during anerobic exercise the excitability of muscle membrane in type I1 fibers becomes more reduced than in type I fibers. Generally, the intramuscular accumulation of lactic
MUSCLE & NERVE
September 1991
835
acid is held responsible for the slowing of MFCV,6,8,z0,2zwhich is in accordance with the findings from IFET during which patients who possess type I1 fibers, indeed, produce more lactic acid. It remains uncertain whether the MFCV of type I fibers will decrease in normal muscle where these fibers are surrounded by lactic acidproducing type I1 fibers. (Figure 2) The initial median frequency of the PDS of 95- 100% type I fibers was minimally lowered, and the shift of PDS during fatigue was less compared with controls. None of these findings reached significant limits. T h e somewhat-lowered mean initial Fmed cannot be explained by the volumeconducting characteristics of the subcutaneous fat layer, acting as a low-pass because the mean percentage fat distribution did not differ significantly between patients and controls. Our findings support the earlier observations that muscles with a higher percentage of type I1 fibers generate a higher Fmedz4 and that the Fmed depends on the MFCV.','* Many authors reported on a shift of PDS to lower frequency in muscle fatigue.3,8,12,1636 I n our experiment, we found a definite shift of Fmed to lower frequency in both patients and controls. This shift of PDS was slightly related to the proportion of fiber types. Two major mechanisms are thought to cause this shift: decrease of the MFCV and changing of the motor unit firing statistics, eg, synchronization.8,z3725T h e slightly lowered shift of PDS in 95- 100% type I fibers can be explained by the absence of MFCV declination in these fibers. Some authors reported that the percentage fall
Power Density Frequency Spectrum.
of Fmed always exceeds that of MFCV,6*36while others found a pro ortional relationship between both parameters. 8,1g19,22 Our findings were in accordance with the first group of findings, and suggest once morezo that the MFCV cannot be the only factor determining the shift of PDS, and that the motor unit firing statistics cause the PDS to shift to lower frequencies. We found no evidence that the influence of the statistics was stronger in patients with 95- 100% type I fibers since, in these patients, the mean value of the negative slope of Fmed can be explained mathematically by the absence of MFCV declination when compared with controls. In conclusion, in patients with type I fiber predominance it was shown that type I muscle fibers have a lower force-generating capacity and a lower MFCV compared with controls. The absence of a MFCV declination in type I fibers indicates an unimpaired membrane excitability during ischemic exercise at 80% MVC, which appears related to the lack of a substantial lactic acid formation by type I fibers. Additional support for our findings originates from the results in patients with 20% type I1 fibers who, in general, show intermediate values between controls and 95- 100% type I fibers. Type I muscle fibers fatigue despite the lack of definite PDS and MFCV changes in the surface electromyogram. T h e EMG amplitude findings also indicate no electrophysiological fatigability. Therefore, these findings support the conclusion that muscle fatigue in patients with 95- 100% type I fiber predominance originates beyond the scope of extracellular SEMG measurements at a sarcolemmal level.
REFERENCES 1. Arendt-Nielsen L, Forster A, Mills KR: The relationship between niuscle fiber conduction velocity and force in human vastus lateralis. J Physiol 1984;353:6P. 2. Banker BQ: The congenital myopathies, in Engel AG, Banker BQ (eds): Myology. New York, McGraw-Hill, 1986, pp 1527-1581. 3. Bigland-Ritchie €3, Donovan EF, Roussos CS: Conduction velocity and EMG power density changes in fatigue of sustained maximal efforts. J Appl Physiol 198 1;51: 1300- 1305. 4. Bigland-Ritchie B, Woods JJ: Changes in muscle contractile properties and neural control during human muscular fatigue. Muscle Nerve 1984;7:691-699. 5 . Braakhekke J P , Stegeman DF, Joosten EMG: Increase in median power frequency of the myoelectric signal in pathological fatigue. Electroenceph Clan Neurophysiol 1989;73:151- 156. 6. Broman H, Bilotto G, DeLuca C: Myoelectric signal conduction velocity and spectral parameters: influence of force and time. J Appl Physiol 1985;58: 1428- 1437.
836
Fatigue in Type I Predominance
7. Brooke MH: Congenital (more or less) muscle diseases, in Brooke MH (ed): A Clinician's Vzew of Neuromuscular DiSeuses. Baltimore, Williams & Wilkins, 1986, pp 340-380. 8. DeLuca CJ: Myoelectrical manifestations of localised muscular fatigue in humans. CRC Crit Rev Bioeng 1984; 112 5 1-279. 9. Donselaar Y, Eerbeek 0, Kernell D, Verhey BA: Fibre sizes and histochemical staining characteristics in normal and chronically stimulated fast muscle in cat. J Appl Physiol 1987;382:237-254. 10. Dubowitz V, Brooke MH: Definition of pathological changes, in Dubowitz V and Brooke MH (eds): Muscle biopsy: A Modern Approach. London-Philadelphia-Toronto, Saunders, 1973, pp 74- 102. 11. Durnin JVGA, Womersley J: Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 7 2 years. 5r J Nutr 1974;32:77-97. 12. Eberstein A, Beattie B: Simultaneous measurement of con-
MUSCLE & NERVE
September 1991
duction velocity and EMG power spectrum changes during fatigue. Muscle Nerve 1985;8:768-773. 13. Hatcher DD, Luff AR: Force-velocity properties of fasttwitch and slow-twitch muscles of the' k;tten. J Physiol 1985;367:377-385. 14. Horita T , Ishiko T: Relationships between muscle lactate accumulation and surface EMG activities during isokinetic contractions in man. Eur J Appl Physiol 1987;56: 18-23. 15. Juel C: Muscle action potential propagation velocity changes during activity. Muscle Nerve 1988;11:714-719. 16. Kereshi S, Manzano G , McComas AJ: Impulse conduction velocities in human biceps branchii muscles. Exp Neural 1983;80:652- 662. 17. Koning FL de, Rinkhorst RA, Kauer JMG, Thijssen HOM: Accuracy of an anthropometric estimate of the muscle and bone area in a transversal cross-section of the arm. Int J Sports Med 1986;7:246-249. 18. K.sler F, Lange F, Caffier G, Kuchler G: Changes of the conduction velocity of isolated muscles induced by altered external potassium concentration. Biomed Biochim Acta 1989;48:465- 470. 19. Lindstrdm L, Peterson I: Power spectrum of EMG signals and its applications, in Desmedt JE (ed): Progress in Clinicul Neurophysiology: Computer-aided Elrctromyopaphy, vol 10. Basel, Karger, 1983, pp 1-51. 20. Linssen WHJP, Jacobs M, Stegeman DF, et al: Fatigue in McArdle's disease. Surface EMG frequency spectrum and muscle fiber conduction velocity during ischemic exercise. Bruin 1990;113:1779- 1793. 21. Maughan RJ, Nimmo MA: T h e inHuence of variations in muscle fibre composition on muscle strength and cross-sectional area in untrained males. J Physiol 1984;351:29931 1. 22. Merletti R, Sabbahi MA, DeLuca CJ: Median frequency of the myoelectric signal. Effects of muscle ischemia and cooling. EurJ Appl Physiol 1984;52:258-265. 23. Miller RG, Giannini D, Milner-Brown HS, et al: Effects of fatiguing exercise on high energy phosphates, force, and EMG: evidence of three phases of recovery. Muscle Nerve 1987; 10:8 10-8'21.
Fatigue in Type I Predominance
24. Milner-Brown HS, Mellenthin M, Miller RG: Quantifying human muscle strength, endurance and fatigue. Arch P I y Mrd Rehabil 1986;67:530-535. 25. Mills KR, Edwards RHT: Muscle fatigue in myophosphorylase deficiency; power spectral analysis of the electromyogram. Electroenceph Clin Neurophysiol 1984;57:330-335. 26. Moritani T, Tanaka H, Yoshida T, et al: Relationship between myoelectric signals and blood lactate during incremental forearm exercise. Am,]Phys Med 1984;63: 122- 132. 27. Ruegg JC: Excitation-contraction coupling in fast- and slow-twitch muscle fibers. Int J Sports Med 1987;8:360-364. 28. Ruff RL: Calcium sensitivity of fast- and slow-twitch human muscle fibers. Muscle Nerve 1989;12:32-37. 29. Sadoyama T, Masuda T, Miyata H, Katsuda S: Fibre cow duction velocity and fibre composition in human vastus lateralis. Eur J Appl Physiol 1988;57:767-771. 30. Sale DG: Influence of exercise and training on motor unit activation. Exerc Sporl Sci Rev 1987;15:95- 151. 31. Schantz P, Randall-Fox E, Hutchison W, Tyden A, Astrand PO: Muscle fibre type distribution, muscle cross-sectional area and maximal voluntary strength in humans. Acta Physiol Scand 1983;117:219-226. 32. Sinkeler SPT, Wevers RA, Joosten EMG, et al: Improvement of screening in exertional myalgia with a standardized ischemic forearm test. Muscle Neiur 1986;9:731-737. 33. Sinkeler SPT: Myoadenylaat deaminase deficiency. Thesis, Nijmegen, T h e Netherlands, 1987, pp 59-70. 34. Yoneda T, Oishi K, Fujikura S, Ishida A: Recruitment threshold force and its changing type of motor units during voluntary contraction at various speeds in man. Bruin Res 1986;372:89-94. 35. Young A: T h e relative isometric strength of type I and type 2 muscle fibres in the human quadriceps. Clin Physiol 1984;4:23-32. 36. Zwarts MJ, Van Weerden TW, Haenen HTM: Relation between average muscle fibre conduction velocity and EMG power spectra during isometric contraction, recovery and applied ischemia. Eur J Appl Physiol 1987;59:212-216.
MUSCLE & NERVE
September 1991
837
