The Effect of Unilateral Eccentric Weight Training and Detraining on Joint ~ngle Specificity, Cross-Training, and the Bilateral Deficit

Transcription

1 The Effect of Unilateral Eccentric Weight Training and Detraining on Joint ~ngle Specificity, Cross-Training, and the Bilateral Deficit )oseph P. Weir, PhD1 Dona ). Housh, Ph D * Terry ). Housh, P ~D loree I. Weir, Ph D4 number of investigations have indicated that isometric strength training results in strength increases that are specific to joint angles at or near the training angle (13.21,!+I,%). Training isometrically at one joint position does not result in strength increases at all points in the range of motion. In addition, a cross-training effect has been shown to occur in which exercise of one limb results in significant increases in strength of the untrained limb (7,8,15,19,24). Recently, Weir et al (24) reported joint angle specificity in both the trained and untrained limbs following isometric strength training. The effects of joint angle specificity and cross-training may have clinical relevance for prescribing resistance exercise in strength rehabilitation. This is especially true in situations where full range of motion exercise and/or bilateral exercise are not possible or contraindicated. Eccentric resistance training is becoming more widely used in rehabilitation (1). Previous investigations have examined various aspects of eccentric resistance training (3,4,6). To date, however, no investigations have Eccentric resistance training is an important component of many rehabilitation protocols. The adaptations following eccentric training are poorly understood in relation to concentric training. The purpose of this investigation was to examine the effects of unilateral eccentric leg extension weight training and detraining on joint angle specificity, cross-training, and the bilateral deficit. Seventeen males volunteered to be subjects for this investigation and were divided into an eccentric training group (N = 91 and a control group (N = 8). The eccentric group performed 8 weeks of unilateral eccentric weight training with the nondominant limb three times per week (3-5 sets of six repetitions) followed by 8 weeks of detraining. These subjects were tested pretraining, posttraining, and following detraining for maximal isometric strength at three joint angles (15, 45, and 75") in both limbs as well as for the one-repetition maximum (I-RM1 eccentric strength of the trained limb, untrained limb, and bilaterally. The results of this investigation indicated that the effects of the eccentric weight training were joint angle specific [significant increases at 45 and 75" (p < This effect was found in both limbs, indicating a cross-training effect that was joint angle specific. The results from the I-RM analyses indicated that the bilateral deficit exists for eccentric contractions (untrained limb > bilateral at pretraining) and that unilateral eccentric training increases this effect (trained and untrained limbs > bilateral posttraining); however, the unilateral training also resulted in increased bilateral strength. Both the I-RM and isometric analyses showed that the training effects persisted over 8 weeks of detraining. Key Words: muscle strength, eccentric training, knee ' Assistant Professor, Program in Physical Therapy, University of Osteopathic Medicine and Health Sciences, 3200 Grand Avenue, Des Moines, /A Assistant Professor, Department of Oral Biology, College of Dentistry, University of Nebraska Medical Center, Lincoln, NE ' Associate Professor, Center for Youth Fitness and Spom Research, Department of Health and Human Performance, University of Nebraska-Lincoln, Lincoln, NE Adjunct Assistant Professor, Program in Physical Therapy, University of Osteopathic Medicine and Health Sciences, Des Moines, IA examined the effects of eccentriconly weight training and detraining on joint angle specificity for isometric strength or cross-training. During dynamic exercise, the biomechanical differences associated with various training devices and limb movements may result in more resistance being placed on the musculature at certain points in the range of motion than JOSFT Volume 22 Number 5 November 1995

2 RESEARCH STUDY others. Thus, it is possible that unilateral eccentric resistance training may result in joint angle specific strength gains that are manifested in both the trained and untrained limbs. Furthermore, the effects of eccentric weight training on joint angle specificity and cross-training following a detraining period have vet to be explored. In addition, several investigations have reported that a bilateral deficit exists in which bilateral strength is less than the sum of the right and left limbs (14). The bilateral deficit has been shown to exist in the forearm flexors and extensors (18) and leg extensors (1 1,14). We are unaware, however, of any investigation that has examined the bilateral deficit under eccentric loading conditions. Therefore, the purpose of this study was to examine the effects of unilateral eccentric weight training and detraining on joint angle specificity, cross-training. and the bilateral deficit. METHODS Subjects Seventeen males volunteered to be subjects for this investigation. The subjects were divided into an eccentric training group (N = 9; % age + SD = years; bodv weight = kg) and a control group (N = 8;?( age + SD = years; body weight = kg). All procedures were approved by the Institutional Review Board and informed consent was obtained prior to any testing. All subjects were physically active but none were engaged in strength training of the lower extremities fbr the 6 months prior to the study. Testing Protocol The eccentric group was tested pretraining, posttraining, and following a detraining period for maximal unilateral isometric strength at three joint angles in both limbs as well as for the one-repetition maximum (1- RM) eccentric strength of each limb individually and bilaterally. The control group was tested at analogous time periods but performed no training. One control subject did not complete all the eccentric I-RM tests and was removed from the 1-RM analyses. The isometric strength tests were performed on a calibrated Cybex I1 isokinetic dynamometer (Cybex, Division of Lumex, Ronkonkoma, NY) at joint angles in which the Cybex I1 lever arm was positioned at 15, 45, and 75" below the horizontal plane. The subject. were stabilized at the torso to the backrest of the device and were further stabilized by holding the handles at the side of the device. Each subject performed two Short-term periods of detraining did not result in losses of strength to pretraining levels. maximal isometric contractions of 3 seconds duration at each joint angle. The contraction with the highest torque plateau was taken as the rep resentative score. A minimum of 2 minutes of rest was allowed between each contraction. The order of testing for the limbs and joint angles was randomized during the pretraining test session. The pretraining order of testing for each subject was also followed during the posttraining and detraining test sessions. High testretest reliability for our isometric procedure has been reported previously (26) Eccentric 1-RM strength was determined on a plate-loaded leg extension weight-training machine (Pro- Star Sports Leg Extension Machine, Blue Springs, MO). For these tests, FIGURE 1. Positioning and stabilization of subject on the leg extension machine. the subjects were seated with their torsos strapped to the backrest of the device and were instructed to hold tightly to the sides (Figure 1). The backrest was adjusted so that the anatomical axes of the knees were aligned with the mechanical axis of the device. The resistance was provided by a lever arm in which weight plates could be attached via a pipe which was welded perpendicular to the long axis of the lever arm. Shin pads were attached to the lever arm for placement against the subjects' legs. The shin pads were a fixed distance from the axis of rotation and, therefore, not adjustable; however, positioning was consistent for each subject across all tests such that strength tests across time were not affected by the intersubject differences in lever arm length. Determination of I-RM eccentric strength involved a trial and error procedure in which progressively heavier loads were applied until the subject could not satisfactorilv complete a repetition. Subsequent trials were performed with lighter loads until the 1-RM was determined to within 2.27 kg. Two minutes of rest were allowed between each trial. To perform each eccentric contraction, an investigator lifted the load to a position where the lever arm was slightly above the horizontal plane. The subject then \'olumc 22 Number 5 November JOSlT

3 TABLE 1. Individual limb isometric torque values (Nm). verbally indicated that he had control of the load, at which time the investigator gently released control. Because of the design of the leg extension machine and to avoid knee joint discomfort, it was decided during pilot data collection to define a satisfactory performance as the ability to stop the rotation of the lever arm at a position where the lever arm was < 5" below the horizontal plane. This resulted in a position where the knee was slightly flexed (< 10"). The subject was then required to lower the weight slowly (1-2 seconds) and in control until the lever arm made contact with a horizontal support. The testing was performed on each limb individually and bilaterally. The order of testing of the individual limbs was randomized for the pretraining test session and was maintained for the posttraining and detraining sessions. The bilateral measure was performed following the unilateral measures. To assess testretest reliability for the 1-RM procedures, intraclass correlation coeficients for the control subjects were calculated. For both the nondominant and dominant limbs, r = 0.97 while r = 0.96 for the bilateral measure. JOSPT Volume 22 Number 5 November 1995 Training Protocol The eccentric group performed 8 weeks of unilateral eccentric weight training on the leg extension device described previously followed by 8 weeks of detraining. The training repetitions were performed as described above. All training was performed with the nondominant limb (as determined by kicking preference) three times per week. Each training session consisted of two to three warm-up sets with progressively heavier loads followed by three to five sets of six repetitions at 80% of the 1-RM. For the first week of training, the subjects performed three sets, the second week the subjects performed four sets, and for weeks three through eight, the subjects performed five sets of six repetitions. The trained limb 1-RM values were retested every 2 weeks to adjust training loads. Statistical Analysis The isometric torque data were analyzed with a four-way [group (eccentric and control) X time (pretraining, posttraining, and detraining) X limb (trained limb and untrained limb) X angle (15, 45, and 75" below the horizontal plane)] mixed factorial analysis of variance (ANOVA), and the eccentric 1-RM strength data were analyzed with a three-way (group X time X limb) mixed factorial ANOVA. To correct for violations of the sphericity assumption, the Huynh-Feldt adjustment was performed when appropriate (12). The limb factor for the eccentric 1-RM analysis had three levels: trained limb, untrained limb, and bilateral. The bilateral I-RM strength score was divided by two in order to make more meaningful comparisons with the unilateral measurements. An alpha level of 0.05 was considered significant for all tests. Isometric Strength The individual isometric limb values across time are presented in Table 1. The analysis showed that there were significant time X angle X limb [F(4,60) = 2.72; p = and time X angle X group [F(4,60) = 3.98; p = interactions (Table 2). Follow-up procedures involved

4 RESEARCH STUDY Effect SS df MS P TXLXA(XG) TXLXA LXA(XG) LX A TXA(XG) TX A TxL(xG) TXL A W ) A G = Time (pretraining, posttraining, and detraining). 1 = limb (trained limb and untrained limb). A = Angle (IS0, 454 and 75'). G = Group (eccentric and control). The p values were determined with Huynh-Feldt corrected df values. TABLE 2. ANOVA table for the four-way isometric analysis. performing three-way ANOVAs (time X limb X angle) for each group (eccentric and control). For the control group analysis, there were no significant (p > 0.05) interactions or main effects for time or limb. As expected, the main effect for angle was significant but was not of interest in this study. For the eccentric group analysis, the three-way (time X limb X angle) and time X limb interactions were not significant (p > 0.05); however, there was a significant time X angle interaction [F(4,32) = 4.44; p = This interaction did not include the limb variable, indicating that there was a cross-training effect following the eccentric weight training. The simple effects for time (collapsed across limb) were analyzed by three (one for each angle) 1 X 3 ANOVA s with time as the indepen- dent variable. The effect for time was nonsignificant (p > 0.05) at 15" but was significant at 45" [F(2,16) = 5.68, p = and 75" [F(2,16) = 8.69; p = Simple comparisons performed with Tukey post hoc procedures within the 45 and 75" models revealed significant increases in is* metric strength (collapsed across limb) at 45" (9.5%) and 75" (13.6%) from pretraining to posttraining. At both angles, the comparisons between posttraining and detraining were not significant. However, at 75", unlike the 45" condition, the mean detraining value decreased somewhat from posttraining such that detraining and pretraining were not significantly different (Figure 2). Eccentric 1 -RM Strength The individual 1-RM limb values for both groups across time are presented in Table 3. The analysis showed that the time X limb X group interaction was significant [F(4,56) = 3.09; p = ; Table 41. This model was subsequently divided into two (eccentric group and control group) two-way (time X limb) repeated measures ANOVAs. For the control group analysis, the time X limb interaction and main effect for time were not significant (p > 0.05). However, the main effect for limb (collapsed across time) was significant [F(2,12) = 10.02; p = and subsequent simple comparisons revealed that the bilateral strength was significantly less than Volume 22 Number 5 November 1995 JOSF'T

5 RESEARCH STUDY model for limb revealed that the untrained limb was stronger than bilateral at pretraining, while both the trained limb and untrained limb were greater than bilateral strength at posttraining and detraining. At no time period were the comparisons between the trained limb and untrained limb significant (Figure 5). DISCUSSION JOINT ANGLE (DEGREES) FIGURE 2. Isometric toque (i t SD) across time for the eccentric group (collapsed across limb); * indicates significantly greater than pretraining (Tukey post hoc analyses). the strength of either the nondominant (p = ) or the dominant (p = ) limb (Figure 3). For the eccentric group analysis, the time X limb interaction was significant [F(4, 32) = 3.13; p = and analyses of the simple effects of time at each limb and of limb for each time period were performed. The simple effects of time were significant for the trained limb [F(2.16) = 13.44; p = , the untrained limb [F(2,16) = 7.01; p = , and bilateral strength [F(2,16) = 17.32; p = Simple comparisons performed with the Tukey procedure Croup within each simple effect model showed that within all limbs, both the posttraining and detraining 1-RM values were significantly higher than the pretraining values; however, none of the posttraining vs. detraining comparisons were significantly different, indicating that the training effect persisted over the detraining period (Figure 4). Similarly, the simple effects of limb were significant at pretraining [F(2,16) = 8.1 1; p = , posttraining [F(2,16) = 12.65; p = , and detraining [F(2,16) = 39.38; p < Tukey post hoc analyses within each simple effect The training in this investigation resulted in a significant time x angle interaction for the isometric analyses in the eccentric group. The effect of the training across time was different at the three joint angles. Follow-up analyses showed significant isometric torque increases at joint angles where the Cybex I1 lever arm was 45 and 75" below the horizontal plane, but the statistically significant effects did not occur at 15". Thus, the effect.. of the eccentric training were joint angle specific. It should be noted that while the simple effect for time was not significant at 15", the changes in torque at 15" from pre- to post-training were 8.8% (collapsed across limb). In the statistical analysis, the effect for limb was not a factor in the eccentric group interactions, indicating the presence of a cross-training effect in which the pattern of increases in isometric strength was similar in both limbs at the angles where significant increases were found (45 and 75"). In this investigation, the increases in isometric strength were Pretraining Posttraining Detraining x SD x SD x SD Eccentric Trained limb Untrained limb O Bilateral Control Nondominant limb Dominant limb Bilateral O 9.1 TABLE 3. Individual limb eccentric I-RM values (kg); I-RM = One-repetition maximum. JOSFT Volume 22 Number 5 November 1995

6 RESEARCH STUDY - - TxL(xG) TXL T(XG) T G T = Time (pretraining, posttraining, and detraining). L = Limb (trained limb, untrained limb, and bilateral). G = Group (eccentric and control). The p values were determined with Huynh-Feldt corrected df values. TABLE 4. ANOVA table for the three-way I-RM analysis; 1 RM = One-repetition maximum. found in both limbs at 45 and 75". The individual limb increases in torque at 45" in the trained and untrained limbs were 15.2% and 4.1%, respectively. At 75", the trained and untrained limb values increased 15.0% and 12.2%, respectively. Following the detraining period, there were no statistically significant decreases in torque from the posttraining values. However, the mean detraining value at 75" decreased somewhat from posttraining so that the detraining and pretraining values were not significantly different. When combined with the observations at 45", in which detraining values were quite close to the posttraining values and were significantly greater than pretraining, it appears that the training effects at 45" were more robust than at 75" and further supports the joint angle specificity effect. Previous investigations have shown joint angle specificity following isometric training (13,21,24,25). The eccentric weight training used in this study was performed through the entire range of motion. Thus, it was expected that strength increases would occur throughout the range of motion. The data analyses, however, indicated that joint angle specificity for isometric strength can occur following eccentric dynamic strength training. Furthermore, the increases in torque at 45 and 75" were evident in both limbs, indicating a cross-training effect that was joint angle specific. These results are similar to the data reported by Weir et al (24) following isometric training. The presence of a cross-training effect is further supported by the significant increases in 1-RM performance in the untrained limb and bilaterally. It is interesting to note that the pattern of joint angle specificity re- ported here is the opposite of what would be expected. The training device, a lever arm which rotates about an axis, would provide the greatest resistance in the horizontal plane as the moment arm for the torque created by the weights is the greatest at that point (23). As the weights are lowered, the moment arm decreases and the resistance torque decreases also (23). Thus, it should be expected that the greatest training effects would occur at joint angles DOMINANT NON-DOMINANT BILATERAL LIMB FIGURE 3. Eccentric one-repetition maximum (I-RM) strength (i + SO) across limbs for the control group (collapsed across rime); * indicates significantly greater than bilateral. Volume 22 Number 5 November 1995 JOSPT

7 RESEARCH STUDY -.. Motor units in a given anatomical locajion may experience a greater training stimulus and, therefore, hypertrophy to a greater degree than other motor units in the muscle. TRAINED UNTRAINED BILATERAL LlMB FIGURE 4. Eccentric one-re~etition maximum (1-RM) strength (i 2 SD) for the eccentric group; * indicates significantly greater than pr&aining. closer to full extension. However, the opposite was found. The effect for time at 15" was not significant while the largest pre- to post-training percent increases were found at 75" (13.6%). The mechanisms underlying joint angle specificity and cross-training are unclear. It has been suggested that adaptations in the nervous system mediate these effects (13,21). However, recent electromyopphic (EMG) studies have shown a dissociation between changes in isometric strength and changes in EMG during maximal contractions of the trained and untrained limbs (24.25). It has been suggested that a specific pattern of hypertrophy may occur following resistance training in which motor units in a given anatomical location may experience a greater training stimulus and, therefore, hypertrophy to a greater degree than other motor units in the muscle (10,17,25). This hypertrophy pattern may then result in increases in strength at specific points in the range of motion. However, this does not explain the angle- specific cross-training effect, as previous EMG studies have shown that the untrained limb is minimally active during unilateral strength training (5,8), indicating an insuflicient stimulus for hypertrophy in the untrained limb. Additionally, statistically significant increases in muscle cross-sectional area have not been reported in the untrained limb (10). Alternatively, these effects may be due to decreases in cocontraction (2). in which training of the quadriceps causes a reduction in hamstring activity during leg extension. The joint angle specificity and cross-training effects in TRAINED UNTRAINED BILATERAL LIMB LIMB PRETRAINING POSTTRAINING DETRAINING FIGURE 5. Eccentric one-repetition maximum (I-RM; i 5 SD) interlimb comparisons for the eccentric group; * indicates significantly greater than bilateral. JOSPT Volume 22 Number 5 November 1995

8 RESEARCH STUDY this investigation may be at least partially explained if decreases in cocontraction were specific to certain angles in the range of motion and occurred in both limbs. Further research incorporating measures of physiological cross-sectional area of the involved muscles (16) and measures of cocontraction may reveal more information regarding the mechanisms of joint angle specificity and cross-training. With respect to the detraining data, the statistical analysis did not reveal a significant decrease in either isometric or eccentric 1-RM performance from the posttraining values over the?-week period. These results extend previous investigations examining the detraining responses following resistance training in which shortterm periods of detraining did not result in losses of strength to pretraining levels (4,9,19,20). Furthermore, the results reported here indicate that both joint angle specificity and cross-training persist following 8 weeks of detraining. These detraining effects were evident in the Tukey post hoc comparisons that followed the significant simple effects for time (collapsed across limb) at 45 and 75" (but not at 15"). These comparisons showed no significant differences between detraining and posttraining values. In addition, the cross-training response over the detraining period is supported by the eccentric 1-RM data, which indicate that the persistence of the training effect was present in both limbs and in bilateral strength. Similar results have been reported by Shaver (19) for the forearm flexors. These data have implications for rehabilitation as prior resistance training in a limb may attenuate the loss in function associated with disuse following surgery and/or immobilization. The eccentric 1-RM data indicated the presence of a bilateral deficit for eccentric strength. This effect was found for the eccentric group at all time periods as the untrained limb was stronger than bilateral at pretraining and both the trained limb and untrained limb were stronger than bilateral at posttraining and detraining. In addition, the control group exhibited a main effect for limb in which follow-up procedures revealed that the bilateral 1-RM values were significantly less than either the dominant or nondominant limb. These data indicate that the bilateral deficit occurs for eccentric as well as isometric (1 1,14,18) and concentric (22) contractions. The bilateral deficit likely results from greater motor unit recruitment during a unilateral as opposed to a bilateral condition (14,22). In strength trained subjects, the bilateral strength is actually higher than the sum of the unilateral strengths (1 1). Therefore, this phe- Bilateral strength is actually higher than the sum of the unila feral strengths. nomenon appears modifiable with training. The data reported here showed that there was a significantly greater I-RM value for the trained limb vs. bilateral for both posttraining and detraining, which indicates that unilateral training resulted in a transfer of the bilateral deficit to the trained limb which persisted across 8 weeks of detraining. It should also be noted that the bilateral 1-RM strength increased posttraining. Thus, while bilateral training is likely the most beneficial for improving bilateral force production, in situations where only unilateral training is an option, bilateral strength is also facilitated. However, more research is needed using procedures that can measure individual limb strength under bilateral test conditions in order to determine whether bilateral strength increases following unilateral training are due to increased force output in both limbs when contracting simultaneously. SUMMARY The results of this investigation indicated that increases in isometric strength of the trained and untrained limbs following unilateral eccentric weight training were joint angle specific and that these effects were present in both the trained and untrained limbs. The eccentric training also resulted in increases in eccentric 1-RM in the trained limb, untrained limb, and bilaterally. The training effects persisted across 8 weeks of detraining. In addition, the bilateral deficit was shown to be present for eccentric 1-RM strength. JOSpT REFERENCES 1. Albert M: Eccentric Muscle Training In Sports and Orthopaedics, p 1. New York, NY: Churchill Livingstone Inc., Carolan B, Cafarelli E: Adaptations in coactivation after isometric resistance training. ) Appl Physiol 73: , Colliander EB, Tesch PA: Effects of eccentric and concentric muscle actions in resistance training. Acta Physiol Scand l4o:3 1-39, Colliander EB, Tesch PA: Effects of detraining following short term resistance training on eccentric and concentric muscle strength. Acta Physiol Scand l44:23-29, Devine KL, LeVeau BF, Yack H): Electromyographic activity recorded from an unexercised muscle during maximal isometric exercise of the contralateral agonists and antagonists. Phys Ther 6 1: , Dudley GA, Tesch PA, Miller B), Buchanan P: Importance of eccentric actions in performance adaptations to resistance training. Aviat Space Environ Med , Gardner GW: Specificity of strength changes of the exercised and nonexercised limb following isometric training. Res Q 34:98-101, Gregg RA, Mastellone AF, Genten JW: Cross exercise--a review of the literature and study utilizing electromyographic techniques. Am J Phys Med 36: , Hortobagyi T, Houmard )A, Stevenson Volume 22 Number 5 November 1995 JOSPT

9 RESEARCH STUDY JR, Fraser DD, Johns RA, Israel RG: The effects of detraining on power athletes. Med Sci Sports Exerc 25: , Housh Dl, Housh TI, Johnson GO, Chu W: Hypertrophic response to unilateral concentric isokinetic resistance training. ] Appl Physiol73:65-70, Howard JM, Enoka RM: Maximum bilateral contractions are modified by neurally mediated interlimb effects. J Appl Physiol 70: , Keppel G: Design and Analysis. A Researcher's Handbook (2nd Ed), p 472. Englewood Cliffs, NJ: Prentice Hall, Kitai TA, Sale DG: Specificity of joint angle in isometric training. Eur J Appl Ph ysiol58: , Koh TI, Grabiner MD, Clough CA: Bilateral deficit is larger for step than for ramp isometric contractions. J Appl Physiol74: , Moritani T, devries HA: Neural factors versus hypertrophy in the time course of muscle strength gain. Am J Ph ys Med 58:ll5-130, Narici MV, Landoni L, Minetti AE: As- sessment of human knee extensor muscle stress from in vivo physiological cross-sectional area and strength measurements. Eur J Appl Physiol65: , Narici MV, Roi GS, Landoni L, Minetti AE, Ceretelli P: Changes in force, crosssectional area and neural activation during strength training and detraining of the human quadriceps. Eur J Appl Physiol59: , Ohtsuki T: Decrease in human voluntary arm strength induced by simultaneous bilateral exertion. Behav Brain Res 7: , Shaver LG: Cross transfer effects of conditioning and deconditioning on muscular strength. Ergonomics 18:9-16, Staron RS, Leonardi MI, Karapondo DL, Malicky ES, Falkel JE, Hagerman FG, Hikida RS: Strength and skeletal muscle adaptations in heavy-resistance-trained women after detraining and retraining. J Appl Physiol70: , Thepaut-Mathieu C, Van Hoecke J, Ma- ton B: Myoelectrical and mechanical changes linked to length specificity during isometric training. J Appl Physiol64: , Vandervoort AA, Sale DG, Moroz 1: Comparison of motor unit activation during unilateral and bilateral leg extension. J Appl Physiol56:46-5 1, Watkins J: An Introduction to Mechanics of Human Movement, pp Boston: MTP Press Limited, Weir JP, Housh TI, Wagner LL: Electromyographic evaluation of joint and specificity and cross-training following isometric training. J Appl Physiol 77: , Weir JP, Housh TJ, Weir LL, lohnson GO: The effect of a unilateral isometric strength training program on joint angle specificity and cross-training. Eur J Appl Physiol fin press) 26. Weir JP, Wagner LL, Housh TI: Linearity and reliability of the IEMG V. torque relationship for the forearm flexors and leg extensors. Am 1 Phys Med 71 : , 1992 JOSPT Volume 22 Number 5 November 1995