Reliability and Validity of Two Isometric Squat Tests

Download PDF
1 downloads 0 Views 243KB Size Report
Comment
small differences in multijoint isometric strength between sub- jects, or .... each trial. Hip and knee angles were checked before each effort and loud verbal ...
Journal of Strength and Conditioning Research, 2002, 16(2), 298–304 q 2002 National Strength & Conditioning Association

Reliability and Validity of Two Isometric Squat Tests ANTHONY J. BLAZEVICH,1 NICHOLAS GILL,2

AND

ROBERT U. NEWTON3

Department of Sports Science, Brunel University, Osterley Campus, Middlesex TW7 5DU, UK; 2Centre for Sport and Exercise Science, The Waikato Polytechnic, Hamilton, New Zealand; 3The Biomechanics Laboratory, Ball State University, Muncie, Indiana 47306. 1

ABSTRACT The purpose of the present study was first to examine the reliability of isometric squat (IS) and isometric forward hack squat (IFHS) tests to determine if repeated measures on the same subjects yielded reliable results. The second purpose was to examine the relation between isometric and dynamic measures of strength to assess validity. Fourteen male subjects performed maximal IS and IFHS tests on 2 occasions and 1 repetition maximum (1-RM) free-weight squat and forward hack squat (FHS) tests on 1 occasion. The 2 tests were found to be highly reliable (intraclass correlation coefficient [ICC]IS 5 0.97 and ICCIFHS 5 1.00). There was a strong relation between average IS and 1-RM squat performance, and between IFHS and 1-RM FHS performance (rsquat 5 0.77, rFHS 5 0.76; p , 0.01), but a weak relation between squat and FHS test performances (r , 0.55). There was also no difference between observed 1-RM values and those predicted by our regression equations. Errors in predicting 1-RM performance were in the order of 8.5% (standard error of the estimate [SEE] 5 13.8 kg) and 7.3% (SEE 5 19.4 kg) for IS and IFHS respectively. Correlations between isometric and 1-RM tests were not of sufficient size to indicate high validity of the isometric tests. Together the results suggest that IS and IFHS tests could detect small differences in multijoint isometric strength between subjects, or performance changes over time, and that the scores in the isometric tests are well related to 1-RM performance. However, there was a small error when predicting 1-RM performance from isometric performance, and these tests have not been shown to discriminate between small changes in dynamic strength. The weak relation between squat and FHS test performance can be attributed to differences in the movement patterns of the tests.

Key Words: movement pattern, multijoint, predicting Reference Data: Blazevich, A.J., N. Gill, and R.U. Newton. Reliability and validity of two isometric squat tests. J. Strength Cond. Res. 16(2):298–304. 2002.

Introduction

R

esearch studies investigating adaptations to weight training often incorporate the free-weight barbell squat as a dominant training exercise (3, 8, 11, 12, 25, 298

27, 30). Nonetheless, although many authors have suggested that the mode (contraction type: isometric, concentric, eccentric) and movement pattern of strength tests should be similar to those of the training exercises (for reviews see Abernethy et al. and Morrissey et al. [1, 19]), relatively few studies use the 1-RM (1 repetition maximum) squat test to determine strength changes after training (25, 27, 30). Instead, strength changes after squat lift training are often examined by isometric tests (10–13, 30), which may be preferred for their high test–retest reliability (2, 4, 28, 30), relatively simple administration, and reduced risk of injury. The relation between dynamic strength increases and isometric strength is not strong (3, 23). For example, Sale et al. (23) found that isometric knee extension strength did not increase after 19 weeks of leg press training despite muscle hypertrophy occurring over the training period. Such results are possibly due to the different contraction modes of training and testing exercises. However, the weak relation between changes in the isometric and dynamic tests may also be related to their different movement patterns. A large body of evidence suggests that adaptations to resistance training are specific to the movement pattern of the training exercises (1, 21, 24, 28). Thus, if isometric tests of strength are to be used in preference to dynamic tests, it may be important that the body position adopted in the isometric test be identical to the training exercise. No research has determined whether the validity of isometric tests is higher when the movement pattern is similar to the dynamic training task than when it is dissimilar. Given that the free-weight squat lift is commonly used in studies investigating adaptations to resistance training, an isometric squat test (IS) might be a useful alternative to the 1-RM squat. However, because the movement pattern of the squat lift is not similar to movements performed in many sports, the isometric forward hack squat (IFHS) may be used since it allows isometric testing with a movement pattern similar to many sports. The IFHS is an isometric version of the forward hack squat (FHS). The FHS is performed in a

Reliability and Validity of Isometric Tests

299

Figure 1. Subject position for both the isometric squat (IS; A) and forward hack squat (IFHS; B) tests.

semiprone position with the hip, knee, and ankle joints moving through similar ranges of motion to movements performed in many sports requiring a pushing motion (e.g., rugby and American football scrimmaging; Figure 1. The joint ranges of motion for 8 subjects are presented in Figure 2 [5, 6]). The movement pattern can also be favorably compared to those attained in the acceleration phase of sprint running (see Jacobs and van Ingen Schenau [18]). Thus the FHS can be considered movement-specific to a number of sports. Given that dynamic versions of the squat and FHS exercise could be used in training programs, it could also be useful to estimate a subject’s 1-RM to prescribe suitable training loads. To minimize time constraints and the potential injury, isometric tests could be used. However, a good relation between isometric and dynamic measures would first have to be shown. The purpose of the present study was first to examine the reliability of both the IS and IFHS tests to determine if repeated measures on the same subjects yielded reliable results, and second to examine the relation between isometric and 1-RM measures of strength. The IS test was performed with a knee angle of 908 and the IFHS test with a hip angle of 908 so that the subjects were in the lowest position of the movement. It was hypothesized that the isometric force would be best related to dynamic 1-RM at this position since it is here that the lifts are most difficult.

Methods Subjects Fourteen athletic men (age range 5 19–26 years) volunteered to participate in the study. At least 11 subjects were required on the basis of an a priori power

test (7, 9) with an effect size correlation of 0.707 (on the basis of an r2 of 0.50, which indicates statistical generality [3]), a of 0.05, and power of 0.80. All subjects played competitive sport at a recreational or representative level and had been using the free-weight squat lift exercise as part of their training for at least 6 months (mean free-weight squat 5 161.6 6 19.5 kg). The research was approved by the Southern Cross University Human Ethics Committee and subjects signed a statement of informed consent. They were able to withdraw from the study at any time. Testing Subjects performed IS and IFHS tests on 2 occasions at the same time of day 1 week apart. Subjects also performed a 1-RM free-weight barbell squat or FHS test on different testing days so that after 2 weeks each subject had performed both the 1-RM squat and FHS tests once. The order of testing was randomized between subjects to prevent order effects, although isometric tests were always performed before 1-RM tests. All tests followed a warm-up including 5 minutes of moderate intensity running and several warm-up repetitions of the free-weight squat and FHS exercises at increasing intensity. IS. Subjects squatted until the internal knee angle was 908 with a 20 kg bar resting across the shoulders. While in this position, the hip angle was measured and recorded. Subjects then moved to a Smith machine (a squat rack designed to allow the bar to move only in the vertical plane) and squatted with its bar across their shoulders until their hip and knee angles were identical to the barbell squat. Metal stops were then placed on top of the bar to prevent its upward movement. Once bar height was established, subjects per-

300 Blazevich, Gill, and Newton

Figure 2. Internal hip, knee, and ankle joint angles during the concentric phase of a forward hack squat (average of 8 subjects) with weight equal to a 60-kg squat lift. 1 radian 5 57.38.

formed 2 warm-up trials of the IS, one at 60% and one at 80% of their perceived maximum exertion (Figure 1). They then performed 3 maximal isometric efforts lasting 4 seconds with 3 minutes of rest separating each trial. Hip and knee angles were checked before each effort and loud verbal encouragement was given to increase subject motivation. Force produced during the squat was recorded by a force platform (Kistler Intrumente, Winterthur, Switzerland) on which the subjects’ feet were placed during each isometric effort. The position of the feet was recorded for subsequent efforts. Force was sampled at 1000 Hz and stored on computer (IBM-compatible 486 DX) for subsequent analysis. IFHS. The rails along which the sled moves were adjusted to an angle of 398 to the horizontal. Subjects placed 2 feet on the foot platform such that the body

formed a straight line from the head to the ankle while in the standing position. They then lowered the weight until the internal hip angle was 908 and the internal knee angle was 1108 (Figure 1b). This approximated the hip and knee angles during push-off in the acceleration phase of sprint running (18). A metal peg was used to hold the machine in this position for the subsequent maximal isometric contractions. Subjects then lifted the sled slowly until the metal peg stopped its upward movement and hip and knee angles were checked to ensure that they were at 908 and 1108 respectively before the subjects provided 2 warm-up (60% and 80% of perceived maximum) and 3 maximal isometric contractions lasting 4 seconds. Three minutes of rest separated each maximal effort. Force produced during the isometric contraction was sampled at 100 Hz by a load cell (output 5 1.9231 mV/V, hysteresis , 0.02%, model LPS-2KG, Scale Components Pty. Ltd., Archerfield, Australia) placed parallel to the direction of sled movement. The signal was fed into a personal computer (IBM-compatible 486 DX) and data stored for later analysis using a custom program written using AMLAB software (AMLAB Technologies, Lewisham, Australia). 1-RM Squat. 1-RM squat strength (free-weight) was tested by subjects lifting increasingly heavy weights until a weight could not be lifted. Subjects placed their feet with the same stance as for the IS test and stood with a loaded barbell across the shoulders. Subjects then squatted until their internal knee angle was 908 before lifting the weight back to the standing position. The smallest increment in weight between lifts was 5 kg. At least 3 minutes separated each trial. 1-RM FHS. The positions of the subjects’ feet and body, and of the rails on which the sled moved, were identical to the IFHS. Each FHS trial required the subject to lower the weight until his internal knee angle was 1108 before lifting the weight back to the standing position. Subjects attempted to lift incrementally heavier weights until a weight could not be lifted. At least 3 minutes separated each attempt and the smallest increase in weight between successive lifts was 10 kg. Statistical Analyses Change in the mean between testing sessions 1 and 2, typical error (i.e., variance of the change in performance between the 2 testing sessions), Pearson’s product moment correlations, and intraclass correlation coefficients (ICCs) were calculated as outlined by Hopkins (16). After curve-fitting procedures were used to ascertain the linear relations between the data (SPSS v10.0, SPSS Inc., Chicago, IL), validity statistics including Pearson’s correlations and linear regression equations with standard errors of the estimates were calculated. For reliability and validity statistics, 95% confidence intervals (95% CI) were calculated for relevant

Reliability and Validity of Isometric Tests

301

Table 1. Reliability statistics for isometric squat and isometric forward hack squat (FHS) tests. Both tests show high reliability.* FHS

D Mean (N) Typical error Pearson’s r ICC

Squat

Mean

LCI

UCI

Mean

LCI

UCI

221.0 29.9 0.99 1.0

262.7 21.5

20.7 49.4

26.9 68.7 0.97 0.97

242.0 48.7

95.9 116.7

0.99

1.0

0.91

0.99

* D mean 5 change in the mean from test week 1 to test week 2; ICC 5 intraclass correlation coefficient; LCI 5 lower limit of confidence interval (95%); UCI 5 upper limit of confidence interval (95%). Table 2. Pearson’s correlations for test performances. The squat lift was highly correlated with the isometric squat (IS). Forward hack squat (FHS) was highly correlated with the isometric FHS. Lower correlations were found between squat and FHS. Test comparisons IS vs. squat IFHS vs. FHS IS vs. IFHS Squat vs. FHS

r

R2

p-value

0.77 0.76 0.47 0.55

0.61 0.59 0.23 0.30

,0.01 ,0.01 .0.1 .0.05

data. Finally, paired t-tests with Bonferroni correction were used to compare observed and predicted (from regression equations) 1-RM test scores to examine differences between data sets. a was set at 0.1 to reduce the likelihood of type II error (finding no difference between observed and predicted values when a difference existed).

Results Reliability of IS and IFHS Reliability statistics for IS and IFHS are presented in Table 1. For IS, there was a small and nonsignificant increase in force produced in the second testing session (26.9 N or 0.9% of 2,321 N). The reliability of the test was very high, with ICC 5 0.97 and typical error of only 69 N. For IFHS, there was a small and nonsignificant decrease in the force produced in the second testing session (26.9 N or 1.2% of 2,335 N). The test– retest reliability of the test was also very high, with an ICC 5 1.00 and typical error only 30 N. Thus, the 2 isometric tests were very reliable. Validity of Isometric Measures There was a significant relation between the average IS (average of testing week 1 and testing week 2) and 1-RM squat performance, and average IFHS and 1-RM FHS performance (see Table 2). There was, however, a poor correlation between subject performances in IS and IFHS tests and only a moderate correlation be-

Figure 3. Scatterplots of isometric vs. 1-RM test performance. The linear regression equations and R2 values are indicated on the graphs. Almost 60% of the variation in 1RM performance can be accounted for by isometric test scores.

tween 1-RM squat and FHS test performances. Therefore subjects who performed well in the isometric tests also performed well in the dynamic tests but subjects who performed well in the squat tests did not necessarily perform well in the FHS tests. Force produced during isometric contractions was converted to weight in kilograms and compared with individual’s 1-RM lifts. On average, IS lifts were 147% of those on the 1-RM and IFHS lifts were only 89% of the 1-RM. Linear regression equations to predict 1-RM performance from isometric performance are presented in Figure 3. The standard error of the estimate for

302 Blazevich, Gill, and Newton

IS was 13.8 kg (95% CI 5 10.9–18.6 kg) and for IFHS was 19.4 kg (95% CI 5 15.4–26.2 kg). These standard errors represent 8.5% and 7.3% of the average performance in 1-RM squat and FHS respectively. There was no significant difference between predicted and obtained values for the data presented here.

Discussion The reliability of both the IS and IFHS tests were very high (ICC 5 0.97 and 1.00 respectively). These ICCs are similar to those previously reported for isometric (0.85–0.99 [2, 4, 26, 29]) and 1-RM tests (0.92–0.98 [14, 15, 17, 22]). The difference in mean performance (shift in the mean) between repeated test occasions was less than 1.5% of the average performance. For the subjects tested here, therefore, there was little or no difference between performances at each testing occasion despite the complex multijoint coordination required to perform the present tests. They could therefore be used to detect small differences in isometric strength between subjects or changes after intervention. There was also a strong relation between subject scores in the isometric tests and the associated 1-RM tests (rsquat 5 0.77, rFHS 5 0.76; p , 0.01), with over 60% of the variation in 1-RM tests explained by subjects’ isometric performances. Thus, subject scores in the isometric tests were strongly related to their 1-RM scores. Furthermore, there was no significant difference between values predicted by regression equations and those obtained by testing of subjects’ 1-RM. Isometric measures could then be considered good indicators of dynamic performance and the equations in Figure 3 could be used to predict 1-RM values. Given the ease with which these isometric tests can be performed, they could be used for such purposes as predicting dynamic training loads. Nonetheless, the correlations obtained here were less than 0.8 and could not be considered indicative of high validity. R2 values for the correlations between isometric and 1-RM tests suggest that up to 40% of the variance in 1-RM performance could be explained by factors other than isometric performance (see Table 2). Since the body positions adopted for the tests were identical, the unexplained variance could be largely attributed to their different movement modes (i.e., isometric vs. dynamic). Furthermore, although the standard errors of the estimates for the relations between the isometric and 1-RM tests were small (SEEsquat 5 13.8 kg, SEEFHS 5 19.4 kg), they still represented 8.5% and 7.3% of the average 1-RM score for the squat and FHS respectively. Thus there was some error in predicting 1-RM performance from isometric performance using the tests presented here. Although performance in the isometric tests could be used as a good indication of a subject’s 1-RM performance, and this performance could be predicted well from the re-

gression equations, precise estimates of 1-RM performance were not possible. It is likely then that data obtained from the isometric tests presented here could not be used to accurately estimate changes in dynamic performance. Given the moderate validity of the isometric tests for estimating 1-RM performance, some modifications could be made to the tests to improve their validity. One change might be to vary the joint angles at which the test is performed. In the present study, joint angles were chosen such that muscle lengths were long and the forces relatively low. However, Sale (22) suggested that test variability was reduced when measurements were taken at the strongest point in the range of motion. Moreover, Murphy et al. (20) found that the elbow angle in a bench press-specific isometric test affected the relation between isometric and 1-RM strength. The authors indicated that tests should be performed at the joint angle at which peak forces were provided. However, it is unclear whether changing the joint angles adopted for the present isometric tests would improve their validity. Future research should also examine the relation between these isometric tests and their associated 1RM tests by investigating the relation between changes in isometric and dynamic strength after a period of resistance training. Although a highly reliable and valid isometric test should measure performance similarly to its comparable dynamic test, this is not always the case. Baker et al. (3) found that a 27% improvement in 1-RM squat and 9% improvement in isometric leg extension force after squat training were unrelated (r 5 0.16, p . 0.05). This was despite significant correlations (r 5 0.57–0.61) between the variables pre- and posttraining that would have indicated moderate validity. Such results are possibly due to the different contraction modes between training and testing exercises. Nonetheless, the low relation between changes in the isometric and dynamic tests may be related to their different movement pattern. The movement patterns of the isometric tests used in the present study were similar to their 1-RM counterparts, thus minimizing the differences between tests. However, it is still unclear whether changes in IS and IFHS test performance would be related to 1-RM squat and FHS test performance after a period of dynamic resistance training. The results of this study also have implications for the movement pattern specificity of performance tests. There was a weak relation between subjects’ performances in the squat and FHS tests such that subjects who performed well in the squat tests did not necessarily perform well in the FHS tests (risometric 5 0.47, r1RM 5 0.55). Given that the tests involved the same contraction modes (either isometric [isometric tests] or an eccentric phase followed immediately by a concentric phase [1-RM tests]), differences between test perfor-

Reliability and Validity of Isometric Tests

mances could be attributed to their different movement patterns. The principle of movement specificity has been shown extensively by past research (1, 3, 5, 6, 19, 28). In the present study, high force production in one posture was not always complemented by high force production in the alternative posture, suggesting an effect of movement pattern on test performance.

Practical Applications The isometric squat and FHS tests were highly reliable and a strong relation existed between isometric and 1RM performance. The IS and IFHS tests could therefore be used to estimate dynamic strength in freeweight squat and FHS tests and allow easy prediction of training loads. Given their high reliability, they could also be used to examine changes in isometric strength between subjects, or after intervention. However, the validity of the tests was moderate (r , 0.8) and the number of subjects tested here reasonably small (N 5 14). Caution should then be exercised when trying to predict a subject’s precise 1-RM from isometric measures. Furthermore, if the isometric tests were to be used to estimate changes in 1-RM strength after intervention, the number of subjects would have to be larger than if a 1-RM test was used. The increase in subject number would equal 1/R2 (16), which for the IS test is 1/0.59, or 1.7, times the subject number. Increasing subject numbers would increase the power of the tests to nullify the loss of power caused by the moderate relation between the test types (lower validity of the isometric measures). Even then it has not been shown that changes in dynamic performance can be accurately assessed by isometric methods; only a training study where both isometric and dynamic tests were performed would determine this. Finally, subjects who produced high forces in one posture (e.g., squat) did not necessarily produce high forces in the other posture (e.g., FHS). There appeared therefore to be an influence of body position on test performance. Thus, to best detect performance changes with training, or differences between subjects, that test that most closely matches the training movement patterns should be used.

References ABERNETHY, P.J., G. WILSON, AND P. LOGAN. Strength and power assessment: Issues controversies and challenges. Sports Med. 19:401–417. 1995. 2. AGRE, J.C., J.L. MAGNESS, S.Z. HULL, K.C. WRIGTH, T.L. BAXTER, R. PATTERSON, AND L. STRADEL. Strength testing with a portable dynamometer: Reliability for upper and lower extremities. Arch. Phys. Med. Rehab. 68:454–458. 1987. 3. BAKER, D., G. WILSON, AND B. CARLYON. Generality versus specificity: A comparison of dynamic and isometric measures of strength and speed-strength. Eur. J. Appl. Physiol. 68:350–355. 1994. 4. BEMBEN, M.G., B.H. MASSEY, R.A. BOILEAU, AND J.E. MISNER.

5.

6.

7. 8.

9.

10.

11.

12.

13.

14.

15.

16. 17.

18.

19.

20.

1.

21.

22.

23.

24.

303

Reliability of isometric force-time curve parameters for men aged 20 to 79 years. J. Appl. Sport Sci. Res. 6:158–164. 1992. BLAZEVICH, A.J., R.U NEWTON, M. SHARMAN, R. BRONKS, AND N. GILL. Specificity of strength training exercises to the sprint run and vertical jump tests [Abstract]. IOC World Congress in Sports Science. Sydney, Australia, November 1999. BLAZEVICH, A.J., R.U. NEWTON, M. SHARMAN, R. BRONKS, AND N. GILL. Specificity of strength training exercises to the vertical jump and 20 m sprint tests. Australian and New Zealand Society of Biomechanics Conference. Gold Coast, Australia, January 2000. COHEN, J. Statistical Power Analysis for the Behavioral Sciences (2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates, . p. 92. DELECLUSE, C., H. VAN COPPENOLLE, E. WILLEMS, M. VAN LEEMPUTTE, R. DIELS, AND M. GORIS. Influence of high-resistance and high-velocity training on sprint performance. Med. Sci. Sports Exerc. 27:1203–1209. 1995. ERDFELDER, E., F. FAUL, AND A. BUCHNER G. Power: A general power analysis program. Beh. Res. Methods Instr. Comp. 28:1–11. 1996. HAKKINEN, K. Neuromuscular fatigue and recovery in male and female athletes during heavy resistance training. Int. J. Sports Med. 14:53–59. 1993. HAKKINEN, K., AND P.V. KOMI. Electromyographic changes during strength training and detraining. Med. Sci. Sports Exerc. 15: 455–460. 1983. HAKKINEN, K., P.V. KOMI, AND M. ALEN. Effect of explosive type strength training on isometric force- and relaxation-time, electromyographic and muscle fibre characteristics of leg extensor muscles. Acta Physiol. Scand. 125:587–600. 1985. HAKKINEN, K., P.V. KOMI, M. ALEN, AND H. KAUHANEN. EMG, muscle fibre and force production characteristics during a 1 year training period in elite weight lifters. Eur. J. Appl. Physiol. 56:419–427. 1987. HENNESSEY, L.C., AND A.W.S. WATSON. The interference effects of training for strength and endurance simultaneously. J. Strength Cond. Res. 8:12–19. 1994. HOEGER, W.W.K., D.R. HOPKINS, S.L. BARETTE, AND D.F. HALE. Relationship between repetitions and selected percentages of one repetition maximum: A comparison between untrained and trained males and females. J. Appl. Sport Sci. Res. 4:47–54. 1990. HOPKINS, W.G. Sportscience: A new view of statistics. http:// www.sportsci.org/resource/stats/index.html. 2000. HORTOBAGYI, T., F.I. KATCH, AND P.F. LACHANCE. Interrelations among various measures of upper body strength assessed by different contraction modes. Evidence for a general strength development. Eur. J. Appl. Physiol. 58:749–755. 1989. JACOBS, R., AND G.J. VAN INGEN SCHENAU. Intermuscular coordination during a sprint push-off. J. Biomech. 25:953–965. 1992. MORRISSEY, M.C., E.A. HARMAN, AND M.J. JOHNSON. Resistance training modes: Specificity and effectiveness. Med. Sci. Sports. Exerc. 27:648–660. 1995. MURPHY, A.J., G.J. WILSON, J.F. PRYOR, AND R.U. NEWTON. Isometric assessment of muscular function: The effect of joint angle. J. Appl. Biomech. 11:205–215. 1995. RUTHERFORD, O.M., AND D.A. JONES. The role of learning and coordination in strength training. Eur. J. Appl. Physiol. 55:100– 105. 1986. SALE, D.G. Testing strength and power. In: Physiological Testing of the High-Performance Athlete. J.D. MacDougall, H.A. Wenger, and H.J. Green, eds. Champaign, IL: Human Kinetics, 1991. pp. 21–103. SALE, D.G., J.E. MARTIN, AND D.E. MOROZ. Hypertrophy without increased isometric strength after weight training. Eur. J. Appl. Physiol. 64:51–55. 1992. THEPAU-MATHIEU, C., J. HOECKE, AND B. MATON. Myoelectrical

304 Blazevich, Gill, and Newton and mechanical changes linked to immobilized muscle. J. Appl. Physiol. 64:1500–1505. 1988. 25. THORSTENSSON, A., J. KARLSSON, J.H.T. BIITASALO, P. LUHTANEN, AND P.V. KOMI. Effect of strength training on EMG of human skeletal muscle. Acta Physiol. Scand. 98:232–236. 1976. 26. VIITASALO, J.T., K. HAKKINEN, AND P.V. KOMI. Isometric and dynamic force production and muscle fibre composition in man. J. Hum. Mov. Stud. 7:199–209. 1981. 27. WILLOUGHBY, D.S. The effects of mesocycle-length weight training programs involving periodization and partially equated volumes on upper and lower body strength. J. Strength Cond. Res. 7:2–8. 1993. 28. WILSON, G.J., A.J. MURPHY, AND A. WALSHE. The specificity of

strength training: The effect of posture. Eur. J. Appl. Physiol. 73: 346–352. 1996. 29. WILSON, G.J., R.U. NEWTON, A.J. MURPHY, AND B.J. HUMPHRIES. The optimal training load for the development of dynamic athletic performance. Med. Sci. Sports Exerc. 25:1279–1286. 1993. 30. YOUNG, W., AND G.E. BILBY. The effect of voluntary effort to influence speed of contraction on strength, muscular power, and hypertrophy development. J. Strength Cond. Res. 7:172–178. 1993.

Address correspondence to Anthony J. Blazevich, [email protected].