Effects of Resistance Training on Elbow Flexors of Highly Competitive Bodybuilders

Effects of resistance training on elbow flexors of highly competitive bodybuilders

STEPHEN E. ALWAY, WALTER H. GRUMBT, JAMES STRAY-GUNDERSEN, AND WILLIAM J. GONYEA Departments of Cell Biology and Neuroscience, and Orthopedic Surgery, University of Texas Southwestern/St. Paul Human Performance Center, University of Texas Southwestern Medical Center at Dallas, Dallas, Texas 75235

ALWAY, STEPHENE., WALTER H. GRUMBT,JAMESSTFUY- Empirical examination of bodybuilders, however, sug- GUNDERSEN,AND WILLIAM J. GONYEA. Effects of resistance gests that women may be capable of substantial increases training on elbow flexors of highly competitive bodybuilders. J. in muscle mass. This idea is supported from our previous Appl. Physiol. 72(4): 1512-1521, 1992.-The influence of work (7,8) in which both average type I and type II fiber genderon muscular adaptation of the elbow flexors to 24 wk of areas, as well as total fiber number, were greater in resis- heavy resistancetraining was studied in five male bodybuilders tance-trained women than in values reported in the liter- (MB) and five female bodybuilders (FB) who were highly competitive. Muscle cross-sectional area (CSA), fiber area, and ature for untrained women (22). In addition, recent lon- fiber number were determined from the bicepsbrachii, and vol- gitudinal data have demonstrated increases in muscle untary elbowflexor torque was obtained at velocities of contrac- mass and fiber area in the quadriceps muscles of women tion between 0 and 3OO”/s. Biceps and flexor CSA was 75.8 and after resistance training (25). Thus it now appears ap- 81% greater, respectively, in MB than in FB, but muscle CSA propriate to conclude that skeletal muscle hypertrophy was not significantly altered by the training program in either in women is possible. group. Becauseestimated fiber number and the volume density It is not clear whether type I and type II fibers in men of nonmuscle tissue were similar in MB and FB, most of the and women respond in identical manners to a similar gender difference in muscle CSA appeared to be due to greater relative load. We have previously determined that the absolute mean fiber areas in MB (10.51 and 10.68 X IO3 pm2 type II-to-type I area ratio in the biceps brachii of women pre- and posttraining, respectively) than in FB (5.33 and 5.96 X lo3 pm2pre- and posttraining, respectively). In neither MB nor bodybuilders was 1.1, and this was significantly less than FB did fiber type achieve further hypertrophy during the 24-wk the ratio of 1.5 that was found in men (7). This suggested training program. These data suggest that the extent of any that female bodybuilders (FB) were unable to achieve a changein musclemass or muscle fiber characteristics is mini- similar degree of preferential hypertrophy in type ‘II mal after a bodybuilder of either gender has attained a high fibers, relative to males. This apparent sexual dimor- degreeof muscle massand a highly competitive status. phism was despite similar training experience and similar relative resistance loads in men and women. How- musclecross-sectional area; type II fibers; fiber cross-sectional ever, because these were cross-sectional data, we could area; fiber number; gender not be certain that women bodybuilders would not dem- onstrate selective type II hypertrophy if their training programs were of similar intensity to those of men. This SEVERAL INVESTIGATIONS have characterized adapta- was considered likely because Staron and co-workers tions of skeletal muscle to heavy resistance training in (26) found that although all fiber types hypertrophied in men (4-6, 19, 22, 27;28), but relatively few studies have the vastus lateralis of women after 20 wk of resistance examined similar responses in women. The studies that training, type II fibers were preferentially increased in have examined the effects of resistance training in their subjects. In a subsequent paper, Staron et al. (25) women are largely limited by methodological weak- observed that 6 wk of resistance training induced prefer- nesses, such as insufficient training duration and inten- ential hypertrophy of type II fibers after a 30- to 32-wk sity. Furthermore, most of such investigations have used detraining period. indirect methods to predict changes in muscle mass (i.e., The purpose of the present study was to determine the percent body fat, limb girth), but these measures do not effects of resistance training on characteristics of type I provide information about changes that might occur at and type II muscle fibers and muscle cross-sectional area the cellular level. Despite these shortcomings, these stud- (CSA). Specifically, our objective was to determine ies have been interpreted by many to indicate that whether type II fibers would respond similarly in men women are capable of significant strength gains with and women to a training program that comprised identi- minimal or no increase in muscle mass (11,30). Presum- cal exercises, sets, repetitions, and relative loads. Al- ably, increases in muscle strength in women would occur though Hakkinen et al. (14,X) have shown that muscle by improvements in neural rather than peripheral mecha- size and strength did not change over a 2-yr period in a nisms (20, 21, 30). group of highly competitive weight lifters, it is not known 1512 0161-7567/92 $2.00 Copyright 0 1992 the American Physiological Society RESISTANCE TRAINING AND MUSCLE 1513 whether muscle mass and strength in highly competitive at high levels of competition in most sports, all of the bodybuilders would show a similar resistance to improve- bodybuilders in this study were interviewed regarding ment. Therefore the second purpose of this study was to their steroid usage. Three of the MB and two of the FB determine whether highly competitive male bodybuilders admitted to current steroid use. The remaining athletes (MB) and FB could further enhance muscle strength, (2 MB and 3 FB) had not taken anabolic steroids for a fiber size, and muscle CSA during a heavy resistance minimum of 4 yr preceding the current study. Blood sam- training period that lasted 24 wk. By training the arm ples were obtained from all athletes, and high-density flexors in highly competitive bodybuilders of both and low-density lipoprotein levels were determined (data genders who had already achieved substantial muscle not shown). Although not all steroids will lower the ratio hypertrophy and strength, we were able to largely elimi- of low- to high-density lipoprotein, our data were consis- nate low motivational factors that may potentially con- tent with the information of current steroid use provided foun .d the interpretation of training studies that have ex- to us by the athletes (i.e., steroid users had very low levels amined previously sedentary subjects. We selected body- of high-density lipoprotein). Because it was not our in- builders- because-their goal is largely to obtain a high tent to manipulate steroid usage but rather to assess the degree of muscle mass in all muscles but especially in the effects of training on the muscle characteristics of these arm flexors. Because bodybuilders train very differently athletes, we excluded subjects after their original inter- from power lifters or weight lifters, we have examined view if they indicated that they would alter their levels only bodybuilders. This eliminated the confounding and/or types of steroids during the course of the experi- problem of other studies that have combined power ment. High- and low-density lipoproteins were not signif- lifters, weight lifters, and/or bodybuilders in a single icantly altered in assessments at 12 or 24 wk of this subject group. study. Therefore we assume that compliance with steroid usage or nonusage was maintained by all subjects. METHODS All women athletes were premenopausal. All had menstrual cycles, although the two women who were using Subjects. Subjects consisted of five MB, five FB, two anabolic steroids reported irregular menstrual cycles. untrained male controls, and two untrained female con- Physical characteristics of the athletes were unaltered trols. Before participating, all subjects signed an in- over the course of the study (Table 1). formed consent form that was approved by the Human Review Board of University of Texas, Southwestern Med- Competitive bodybuilders, like other athletes, train differently during the “off-season” than during the “com- ical Center at Dallas. Data obtained before training have petitive season.” been reported on some of these subjects in a previous Typically, the off-season is a period study (7). All of these athletes were highly competitive when bodybuilders attempt to improve muscle mass and because they had won their weight or height class at ei- strength in all body parts and especially in those that ther National Physique Co mmittee- or American Ath- might have shown some asymmetry during the previous letic Union-sanctioned bodybuilding championships in competitions. Approximately 8-12 wk before a body- the states of Texas, Arkansas, or Louisiana. Three of the building competition, bodybuilders will reduce their calo- MB and two of the FB had placed second through fifth at ric intake and attempt to reduce subcutaneous body fat national-level bodybuilding championships held by the stores, while maintaining (but not adding to) the existing National Physique Committee or the American Athletic levels of muscle mass. Union. These were non-drug-tested events at the time of The bodybuilders in the current study were exam- data collection. One MB had won a Natural Physique ined 24 wk after their most recent competition (range of Competition national meet where drug testing against 6-16 wk). During the 24 wk of this study, none of the steroid use was enforced. bodybuilders competed in a bodybuilding competition, Dietary patterns for these athletes were assessed be- nor did they attempt to reduce caloric intake to reduce fore and after the study by a questionnaire that recorded body fat stores. Thus they would have been considered in their food intake for 7 consecutive days. The athletes the off-season part of their competitive schedule. recorded the time of food intake, the type of food, and the These athletes trained with resistance exercises for 4 -t weight of each food item. All athletes supplemented their 1.1 h/day, 5-6 days/wk, and no difference was found be- diets with multivitamins, protein powders (from milk tween subject groups. We did not attempt to control and soy products), and free-form amino acid capsules. In training for muscle groups other than the arm flexors, so addition, two MB and one FB used a carbohydrate sup- the bodybuilders trained their other body segments plement before training. Dietary intake averaged 6,422 t (thighs, chest, back) in the normal manner. A question- 877 and 4,241 t 338 (SE) kcal- for MB and FB, respec- naire indicated that the subjects typically used 15-20 sets tively, and this was consumed over 6-8 feedings per day. of exercises for their thigh, chest, and back muscles, The nutrient composition was estimated from- the di- whereas only 12-15 sets of exercises were used for their etary records of the athletes. Nutrient composition was shoulder, triceps, and triceps surae muscles. Likewise, similar among the subject groups, and it consisted of 67 t the history of total length of training did not differ be- 11% carbohydrate, 22 t 9% protein, and 11 t 4% fat. tween subject groups (MB 5.8 t 0.5 yr and FB 5.4 -+ Nutrient composition of the diet was unchanged before 0.7 yr). and after 24 wk of the training study. Training program. The training program for the arm Because steroid use has been common among athletes flexors was designed to reflect the type that is typically 1514 RESISTANCE TRAINING AND MUSCLE

TABLE 1. Subject characteristics

Subj No. Age, yr Height, cm Body Mass, kg FFM, kg FFM/Body Height, kg/cm Training, yr

MB 39 171.4 72.1 68.8 0.40 5.2 35 173.9 82.5 73.4 0.42 4.8 29 176.6 105.2 92.5 0.52 5.5 28 169.5 96.7 86.6 0.51 6.4 33 163.8 91.6 79.9 0.49 7.3 Mean k SE 32.8t2.0 171.3t1.6 89.6t2.3 80.2*5.7* 0.47t0.02* 5.8kO.5 MC 1 26 172.1 74.4 52.1 0.30 2 29 174.8 78.9 62.6 0.36 Mean t SE 27.5t1.5 173.5kl.4 76.7t2.3 57.4t5.3 0.33t0.03 FB 31 168.3 66.5 56.5 0.34 6.8 35 165.1 71.8 61.2 0.37 7.1 34 162.5 62.8 51.5 0.32 4.4 38 172.1 64.4 52.6 0.31 3.9 36 156.8 53.2 42.1 0.27 4.8 Mean 2 SE 34.8t1.2 165.0k2.6 63.7t3.0 52.8t3.2 0.32kO.02 5.4kO.7 FC 1 32 167.6 58.3 47.1 0.28 2 26 171.8 57.7 46.6 0.27 Mean k SE 29.Ok3.0 169.7t2.1 58.OkO.3 46.9t0.3 0.28t0.01 FFM, fat-free mass; MB, male bodybuilders; MC, untrained male controls; FB, female bodybuilders; FC, untrained female controls. * P < 0.05, MB vs. FB. conducted by advanced bodybuilders of both sexes. This and 14 repetitions of each exercise in the second training was determined after interviewing and observing 18 male session. and 15 female bodybuilders of state or national competi- Fat-free mass (FFM). The total body volume of sub- tion experience who performed the appropriate arm jects was determined by water displacement in a body flexor exercises in their normal training routines. The volume tank (Whitmore Enterprises, TX) according to training regimen is summarized in Table 2. It consisted procedures outlined previously (7, 8, 12, 16). Residual of two parts, with each part followed once per week (the lung volume was determined by the method of Wilmore arm flexors received two training sessions per week). A et al. (31). Body volume was calculated as follows: body 6-repetition maximum (RM) and 10 RM were deter- volume = total body volume displacement from volume mined for each of the four exercises. On the first training meter - residual lung volume - abdominal gas volume. session, each exercise was performed for five sets of six Abdominal gas volume was assumed to be 0.11 liter (10). repetitions in each set, with use of the resistance that Mass was determined on land, and density was deter- equaled six RIM. In the second exercise session, the resis- mined according to the following equation: density = tance was reduced so that three sets of 10 repetitions mass f volume. Percent body fat was determined from were completed with the resistance that achieved 10 RM. the equation of Siri (23). FFM was calculated as total The resistance was increased for an exercise when they body mass - (total body mass X percent fat). Six subjects could perform 8 repetitions in the first training session were brought in to determine the coefficient of variation from test to test, which was 3.2 t 0.3% on these subjects. TABLE 2. Exercise training program for MB and FB Percent fiber distribution. Two needle biopsies were obtained from the same site from the long head of the bi- Program 1 Program 2 ceps brachii before and after 24 wk of heavy resistance training. The depth of biopsy was controlled so that the Exercise Sets Reps Sets Reps same region of the muscle was sampled at both intervals. Barbell curls 5 6 3 10 The tissue was frozen in isopentane cooled to the temper- Standing alternate dumbbell curls 5 6 3 10 ature of liquid nitrogen and was stored at -7OOC before Barbell Scott curls 5 6 3 10 processing for histochemistry. Cryostat sections were cut Bent-over dumbbell concentration curls 5 6 3 10 at pm, mounted on glass slides, and stained for myofi- Hammer curls 5 6 3 10 10 brillar adenosinetriphosphatase (mATPase) activity (3, The training regimen was alternated between programs 1 and 2 each 7,9). Fibers were classified as type I or type II from the week. Program 1 used a 6 and 2 a 10 repetition maximum (RM) resistance. Two minutes of rest occurred between each set in program 1 and mATPase histochemical reactions after pH preincuba- 60 s of rest occurred between sets in program 2. MB, male bodybuilders; tions of 3.29 or 10.25, respectively. Photographic mon- FB, female bodybuilders. tages were assembled, all the fibers in the biopsy sample RESISTANCE TRAINING AND MUSCLE 1515 were counted (667 t 73), and percent fiber distribution that if the degree of sarcomere shortening is the same in was determined. both subject groups and before and after training, valid Fiber CSA. Fiber areas were determined by planimetry intergroup comparisons can still be made on fiber num- on a minimum of 200 type I and 200 type II fibers of each bers that are either corrected or uncorrected for sarco- biopsy, because we have previously found that samples mere shortening. fewer than this may result in errors in fiber area data for Isometric torque. Voluntary isometric torque was de- bodybuilders (7). For this reason, fiber areas were deter- termined before and after 12 and 24 wk of heavy resis- mined for only the major fiber types, namely type I and tance training. The Cybex II dynamometer (Ronkon- type II fibers. Light micrographs were taken (X100) of koma, NY) was set at O”/s to record isometric torque. cross sections of fibers after an alkaline mATPase reac- Preliminary studies on four male and four female body- tion (pH 10.25) and were printed to a final magnification builders determined that an elbow angle of 90’ (straight of about X15,000. Fiber perimeters were manually traced arm 180”) produced the greatest torque. There did seem on a summagraphics digitizing tablet. A minicomputer to be some gender differences, because torque that was software computer program was used to store individual recorded at elbow angles of 7O,90, and 120’ were similar fiber areas and also to calculate type I and type II fiber in females, whereas in males, isometric torques were simi- area for each subject. Mean type I and type II fiber area lar and the greatest at elbow angles of 90 and 120”, but values were determined for MB and FB groups by averag- these were always less than torques generated at an el- ing fiber area from the respective fiber type for each sub- bow angle of 70°. Intertest variance for isometric torque ject in the sample group. In addition to the areas for type was assessed in six subjects at a 1-wk interval, and the I and type II fibers, mean fiber area (F,) was calculated as coefficient of variation was 4.7 t 0.9%. outlined previously (7). Peak isokinetic torque. Peak torque (PT) was defined Volume density of nonmuscle tissue. Nonmuscle tissue as the greatest torque developed during the contraction was identified on frozen tissue sections stained for Go- after the initial impact torque and was determined with moris’ trichrome (7, 9). The volume density of collagen an isokinetic dynamometer (Cybex II). Impact torque and other nonmuscle tissue was determined from a l21- (13, 32) was not used for calculations of peak flexion or point grid lattice according to standard stereological tech- peak extension torques. Each subject received several niques (29) as previously reported (2, 3, 7, 9). submaximal practice contractions at each velocity to fa- Muscle area. CSA of the biceps brachii and total arm miliarize him with the test protocol. Four to five maximal flexor mass (biceps brachii + brachialis muscles) was de- voluntary efforts were performed at angular velocities of termined from computerized tomographic (CT) scans of 30,60,180,240, and 3000/s. Contractions at each velocity the upper arm before the muscle biopsy (7). The hand were separated by a minimum of 4 min to provide suffi- was supinated, and the arm was in the extended position. cient recovery. At the end of the study, subjects per- Six scans were taken at 3-mm intervals from the mid- formed one to two maximal contractions at 30’ and at belly region of the upper arm containing the greatest area 3OO”/s to check for signs of fatigue. These postsession of the biceps. CT scans were taken with a soft tissue measures were within intertrial variations. We assume algorithm. This computer algorithm was part of the soft- that all of the subjects were contracting maximally at ware package produced by the manufactures of the CT each velocity because the coefficient of variation on scan unit that optimized the visualization of muscle and torque measures did not exceed 5.4 t 1.1% during repeat other soft tissue (Picker International, TX). Window measures of torque that were determined a minimum of 1 and level settings were 160 and 64, respectively. The CSA wk apart. Isokinetic PT measures were obtained before of the biceps brachii was determined by planimetry from training, after 12 wk of training, and after 24 wk of five traces of each scan. The scan that had the greatest training. CSA was selected to represent the biceps CSA. Both Free-weight strength measures. Ten and 6 RM mea- flexor and biceps CSA were corrected for the percentage sures were recorded before, at 12 wk, and after 24 wk of of collagen and noncontractile tissue that was deter- training. These data were used largely to assess the at- mined from biopsies of the biceps. Muscle CSA was tempts of the subjects to improve the resistance for each determined before training and after 12 and 24 wk of exercise. Because the style of lifting (speed of movement, training. hand, elbow, and shoulder position) is not carefully con- Fiber number. Biceps fiber number was estimated as trolled in free-weight assessments of strength, improve- follows: number of fibers = biceps CSA (corrected for ments in strength may not necessarily be the result of noncontractile tissue) + F,. changes in the ability of the muscle to produce force. Muscle obtained by the needle biopsy technique un- Nevertheless, it is recognized that training-specific dergoes almost complete contracture postexcision (18); changes are better assessed on the particular apparatus therefore fiber areas were measured with sarcomeres in a that was used for training. Thus, because the Cybex isoki- contracted state. Because muscle CSA was measured by netic device may not be a sensitive tool for determining CT scanning with the sarcomeres at resting length, the changes due to free-weight exercises, assessments of 10 net result is that F, has been overestimated and therefore and 6 RM were obtained. fiber number has been underestimated. We have there- PT-to-muscZe CSA ratios. Torque is the result of mus- fore corrected the estimates of fiber number by 36% to cle force acting at a muscle moment arm length. An addi- account for sarcomere shortening, according to the cor- tional series of CT scans might have been able to deter- rection of MacDougall and colleagues (18). We assume mine the distance from the insertion of the bicipital ten- 1516 RESISTANCE TRAINING AND MUSCLE

I - 24 week training study - 1 the two subject groups. FFM was, however, significantly 6-16 wks 0 12 24 8-16 wks greater in MB. FFM did not change in either group over I I-Ir-----r-I------l the duration of the study. Bodybui lding Biopsy CT scan B i opsy Bodybui lding Competition CT scan Strength measures CT scan Competition Elbow flexor and biceps CSA. Flexor and biceps CSA Strength measures Body coqmsi tion Strength measures Body composition Body composition are given for each subject in Table 3. Because CT scans FIG. 1. Schedule for data collection during X-wk training study. were not obtained from one FB at week 24, data in Table Pretrained data were collected 26 wk after last bodybuilding competi- 3 are given both for the entire group (n = 5, data in tion. Bodybuilders did not compete in next competition until 28 wk brackets) and for the four FB who completed the entire after completion of study. study. Biceps and flexor CSA (biceps brachii + brachialis) were 24.8 t 2.6 and 34.4 t 2.8 cm2, respectively, in don to the point of axis rotation at the elbow, but this MB. Biceps and flexor CSA were 14.1 t 2.1 and 19.0 t 2.9 would have induced unacceptable amounts of radiation cm2 (FB n = 5,19.6 & 1.9 cm2), respectively, in FB. Thus to the subjects. Furthermore, the point of insertion of the MB had 76 and 81% greater CSA in biceps and flexor biceps brachii and brachialis and the origin of the bra- groups than in the FB group. The small increase in flexor chioradialis would have been very difficult to determine muscle CSA of 3.1 t 0.8% in MB and 4.2 t 0.9% in FB from CT scans. As a result, relative PT was determined (FB n = 5, 3.7 t 0.5%) after 12 wk of training was not from the ratio of PT to muscle CSA. In addition, isomet- statistically significant. Similarly, the increase of 7.6 t ric torque and/or CSA was determined. Because an el- 1.5 and 7.8 t 1.1% in MB and FB, respectively (FB n = 5, bow angle of 90’ produced the greatest isometric torque 5.8 +- 0.4%), after 24 wk of training (relative to pretrain- in both males and females, the data for this angle were ing) did not achieve statistical significance (training ef- selected for measurement of ratios of isometric torques fect P = 0.09). Because there were no statistically signifi- to CSA in both subject groups. cant differences in the absolute degree of improvement Muscle CSAIFFM. We (5, 7) have previously found between the subject groups, data from all subjects were strong linear correlations between muscle CSA and pooled. The pooled data demonstrated a significant im- FFM. Therefore subjects may have had large arms be- provement in flexor muscle CSA after 24 wk of training, cause they had greater total FFM. Thus, to normalize for but no changes were found after 12 wk of training. CSA FFM, we calculated flexor CSA/FFM. in the biceps did not change in either subject group or Time course of data collection. All subjects were exam- when the data were pooled for either 12 or 24 wk of train- ined before training, after 12 wk of training, and after 24 ing. The ratio of flexor to biceps CSA was not signifi- wk of training. Biopsies were obtained before and after cantly different between subject groups, and it was not training only (Fig. 1). altered by training (Table 3). Statistics. Descriptive statistics includes means t SE. Flexor CSA/FFM. Because persons with a high FFM The pre- to posttraining changes of all performance and also have a large muscle CSA (7), flexor CSA/FFM was CT scan data were analyzed by a 2 (male/female) X 3 calculated to normalize for body FFM among subjects. (pretraining, midtraining, posttraining) analysis of vari- This calculation demonstrated that muscle CSA was still ance (ANOVA) with repeated measures. Data from the significantly greater in MB than in other groups (Table biopsy (fiber area, fiber type, percent nonmuscle tissue, 3). This relationship was, however, unaffected by the etc.) were analyzed by a 2 (male/female) X 2 (pretraining, training program. posttraining) ANOVA with repeated measures. The in- Isometric torque. Voluntary isometric torque was lower teraction term from the two-way ANOVA was used to at an elbow angle of 70’ than at either 90 or 120’ in MB. determine the difference in adaptation between men and This did not change over the course of the study. Differ- women. Post hoc analyses were performed by the New- ing the elbow angles did not significantly alter develop- man-Keuls procedure. Correlations were analyzed by the ment of isometric torque in FB. Although there was test- Pearson produce moment with no separation of subjects to-test variability in the data, there were no significant by gender. One-way x2 analysis was performed between changes in voluntary isometric torque over 24 wk of subject groups on frequency-area data. P < 0.05 was heavy resistance training in either subject group (Fig. 2). selected to indicate statistical significance. p was set Peak isokinetic torque. MB had significantly greater iso- at 0.10. kinetic torque at all velocities of contraction than FB. PT was greater at l2O”/s than at contraction velocities of RESULTS 180, 240, and 3OO”/s for MB subjects. PT at 6O”/s ex- Subjects. One of the five women athletes became preg- ceeded the torque generated at all other velocities in both nant during the final 4 wk of the study but continued her MB and FB at a given test period (Fig. 3). training. CT scans and isokinetic strength measures Flexor PTICSA. PT was strongly correlated to flexor were not obtained on this subject at week 24 because of CSA (pretraining R = 0.89, P < 0.001 and posttraining potential complications with the pregnancy; however, a R = 0.90, P < 0.001). Thus subjects with large muscles needle biopsy was obtained at this time point. Statistical were able to produce more torque than those with analyses were conducted on muscle CSA and strength for smaller muscles. In an attempt to normalize for torque, a only the four women who completed the study in its en- measure of relative torque was calculated as flexor PT/ tirety; however, because biopsies were obtained on all flexor CSA. Isometric PT/CSA at an elbow angle of 90° five women before and after training, statistical analyses averaged 2.9 t 0.1 and 2.8 t 0.1 in FB and MB groups, for fiber area included all five FB. respectively, before and after 12 and 24 wk of resistance Age, height, and years of training did not differ among training. This ratio was also similar in MB and FB at all RESISTANCE TRAINING AND MUSCLE 1517

TABLE 3. Flexor and biceps CSA

Flexor CSA Biceps CSA Flexor GSA/Biceps CSA Flexor CSA/FFM

Subj No. Owk 12 wk 24 wk Owk 12 wk 24 wk Owk 12 wk 24 wk Owk 12 wk 24 wk

MB 1 31.6 29.4 27.9 21.6 19.6 20.0 1.5 1.5 1.4 0.46 0.43 0.41 2 26.8 28.5 29.1 18.5 18.6 18.7 1.4 1.5 1.6 0.37 0.39 0.40 3 40.1 39.8 43.4 28.6 26.3 31.3 1.4 1.5 1.5 0.43 0.43 0.47 4 31.9 34.8 34.8 22.5 23.2 24.6 1.4 1.5 1.4 0.37 0.40 0.40 5 41.7 45.6 49.7 32.9 32.2 33.8 1.3 1.4 1.5 0.52 0.57 0.62 Mean t SE 34.4t2.8* 35.5t3.1* 37.0t4.1* 24.8t2.6* 24.0t2.5* 25.7t2.9* 1.4tO.l 1.5tO.l 1.5tO.l 0.43t0.03* 0.44t0.03* 0.46t0.04* MC 1 21.3 20.9 21.1 14.4 14.1 14.2 1.5 1.5 1.5 0.41 0.40 0.40 2 22.9 22.6 23.1 15.5 15.2 15.3 1.5 1.5 1.5 0.36 0.36 0.37 Mean t SE 22.1t0.8 21.8t0.9 22.ltl.O 15.0t0.6 14.7t0.6 14.8t0.6 1.5tO.O 1.5tO.O 1.5tO.O 0.39t0.03 0.38&0.02 0.93,tO.O2 FB 1 15.7 17.8 17.7 10.3 12.4 11.3 1.4 1.4 1.5 0.26 0.31 0.31 2 24.1 23.6 24.9 19.0 17.9 16.7 1.4 1.3 1.5 0.44 0.39 0.41 3 14.4 15.9 15.8 11.0 10.4 10.7 1.3 1.5 1.5 0.28 0.31 0.31 4 21.6 22.3 23.8 15.9 16.4 16.5 1.3 1.4 1.4 0.38 0.42 0.45 5 22.2 21.9 16.5 16.7 1.3 1.3 0.52 0.52 Mean t SE 19.0t2.9 19.9t1.8 20.6t2.2 14.1t2.1 14.3t2.1 13.8t1.5 1.4tO.l 1.4*0.1 1.5tO.l 0.34t0.04 0.36t0.03 0.37t0.04 Mean k SET 19.6t1.9 20.3tl.5 14.5t1.7 14.8t1.4 1.3kO.l 1.4kO.l 0.38kO.05 0.39t0.04 FC 1 10.1 10.1 10.0 8.0 8.1 8.1 1.3 1.2 1.2 0.21 0.21 0.21 2 12.1 11.7 11.9 9.7 9.5 9.6 1.2 1.2 1.2 0.26 0.25 0.26 Mean t SE 11.ltl.O 10.9t0.8 ll.Otl.0 8.9t0.9 8.8t0.7 8.920.8 1.3-t-0.1 1.2tO.O 1.2tO.O 0.24kO.03 0.23t0.02 0.24t0.03 CSA, cross-sectional area; CT, computerized tomographic. For other abbreviations, see Table 1. * P < 0.05, MB vs. FB. t Data for 5 FB; no CT scans for 1 FB at 24 wk. tested contraction velocities, but the ratio decreased par- other noncontractile tissue was greater in females than allel to the drop in torque (as a result of the force-velocity in males and averaged 11.1 t 0.5 and 9.0 t 0.2%, respec- relationship). tively. Identical measures were obtained from the biopsy Fiber type distribution. The percentage of type I and obtained before and after 24 wk of training. Because the type II fibers did not differ among subject groups. Type II percentage of nonmsck tissue did not change during fibers averaged 62.1 t 2.6 and 58.0 t 2.9% in MB before the duration of the study in MB and FB, correcting flexor and after training, respectively. The percentage of type II or biceps CSA did not change the relative differences fibers in FB was 50.1 -t 2.2 and 54.9 + 2.8% before and between subject groups. after training, respectively. 100 Nonmuscle tissue. The volume densitv of collagen and - c 1 ** O- 0 pre-training weeks weeks 2 g 60 i MBP B MB 1 & 40 s *** z 20 FB 3 O- 0 pre-training 9 0 I 1 I I 1 l - 0 12 weeks 60 120 180 240 300 q - Cl 24 weeks VELOClIYOFCUNTRACT1ON (o/s) I 7b 90 120 FIG. 3. Voluntary isokinetic torque of elbow flexors before training ELBOW ANGLE ( 0) and after 12 and 24 wk of heavy resistance training in 5 MB and 4 FB. Contraction velocities ranged between 60 and 300*/s. **P < 0.05, volun- FIG. 2. Isometric torque of elbow flexors before training and after tary isokinetic torque was greater at 60*/s than at all other velocities. 12 and 24 wk of heavy resistance training in 5 male bodybuilders (MB) *P < 0.05, torque at contraction velocity of 120*/s was greater than at and 4 female bodybuilders (FB). **P < 0.05, voluntary torque was less faster velocities of contraction. ***P < 0.05, torque generated at 60*/s than torque generated at other elbow angles. Straight arm = elbow was greater than torque generated at velocities of contraction between angle of 180*. Voluntary torque was greater in MB than FB at all elbow 180 and 240*/s. Voluntary isokinetic torque was greater in MB than FB angles. Values are means t SE. at all velocities of contraction. Values are means t SE. 1518 RESISTANCE TRAINING AND MUSCLE

TABLE 4. Muscle fiber characteristics

Type II Area/ Type II Area, X103 pm2 Type I Area, X103 pm2 Mean Fiber Area, X103 pm2 Type I Area

Subj No. Owk 24 wk Owk 24 wk Owk 24 wk Owk 24 wk

MB 1 13.90 12.97 9.60 9.53 12.2 11.8 1.5 1.4 2 7.98 7.96 6.14 5.93 7.47 7.49 1.3 1.3 3 13.60 12.63 8.72 8.39 11.49 10.99 1.6 1.5 4 15.15 15.47 9.00 8.97 12.12 12.05 1.7 1.7 5 10.18 12.47 6.97 8.40 9.29 12.08 1.5 1.6 Mean t SE 12.16tl.33 12.30tl.21 8.09kO.65 8.24t0.62 10.5lt0.93 10.68t0.82 1.5kO.l 1.5tO.l MC 1 4.41 4.12 3.88 3.97 4.09 4.06 1.1 1.0 2 4.75 4.88 4.11 4.34 4.43 4.61 1.1 1.1 Mean k SE 4.58t0.17 4.5OkO.38 3.99t0.12 4.16kO.19 4.2620.17 4.33k0.28 l.lkO.0 1.lt0.1 FB 1 4.72 4.47 4.14 3.83 4.41 4.39 1.1 1.2 2 8.74 8.88 7.98 7.22 7.95 8.23 1.1 1.2 3 5.14 6.88 4.84 5.91 4.98 6.41 1.1 1.2 4 4.85 6.02 4.45 5.86 4.68 5.85 1.1 1.0 5 4.65 5.00 4.60 4.88 4.63 4.91 1.0 1.0 Mean + SE 5.61t0.77 6.25t0.77 5.2OH.11 5.54t0.57 5.33t0.66 5.96t0.67 l.lkO.1 l.lt0.1 FC 1 3.11 2.88 3.01 2.68 3.07 2.81 1.0 1.0 2 3.87 3.05 3.77 3.00 3.83 3.03 1.0 1.0 Mean t SE 3.49t0.38 6.25kO.19 3.39t0.24 2.84kO.17 3.45t0.16 2.9lt0.17 l.Ot0.0 1 .OtO.O For abbreviations, see Table 1.

Fiber area. F, (corrected for fiber type distribution) number over the course of the 24wk training period were averaged 10.51 X lo3 pm2 in MB and 5.33 X lo3 pm2 in FB not significant for either subject group. These small (Table 4). This represented a significantly greater F, in changes in estimated fiber number likely indicate the MB and was the result of increased areas in both type I variability in the estimation procedure. and type II fibers. Type I and type II fiber areas were 56 and 96% greater, respectively, in MB than the corresponding fiber areas of FB, pre- and posttraining. Type II-to-type I fiber area ratios were 1.5 and 1.1 in MB and FB, respectively (P < G Li- 10 O.O5), before and after training. 0 8 Fiber area data plotted as a frequency distribution G z6 curve indicated that average type II fiber area reflected W not only an organization of only larger fibers but also a 14 23 2 wider distribution of type II fiber areas in MB (Fig. 4) z 0 relative to FB. One-way x2 analysis of the type II fiber 6000 10000 14000 2000 6000 10000 14000 data indicated that MB and FB had a similar frequency FIBER AREA (urn *) FIBER AREA (urn ‘) of fibers ~0.5 X lo3 pm2 in area; however, MB had a z8? c D significantly greater frequency of fibers between 0.5 and u-l 10 0 2.0 x lo3 pm2 relative to FB (pretraining 4.1 vs. 1.9%, I5 g W 0 pre-training E8 m 8 posttraining 3.7 vs. 1.3%; P < 0.01). LL 7 m 24 weeks 7 9 I Average type I fiber area reflected a distribution of b 6 k6 fibers larger in MB than in FB (Fig. 4). The frequency of _ t; 5 0F5 type I fibers between 0.5 and 2.0 X lo3 pm2 was less in MB 5 4 z4 than in FB (pretraining 1.5 vs. 3.8%, posttraining 2.1 vs. 3 Y 3 g2 Ei 2 4.3%; P < 0.01). E 1 E ' Fiber number. Estimated fiber number for MB was IA- o,....,....,....,. .., 0 , I 5000 10000 15000 20000 5000 10000 15000 20000 227,829 t 36,637 and 243,771 t 32,016 pre- and post- FIBER AREA (urn* ) FIBER AREA (urn’ ) training, respectively. Fiber number for FB was esti- FIG. 4. Fiber area-frequency distribution curves for type I (A and C) mated to be 239,208 t 19,637 and 232,208 t 18,223 pre- and type II (B and D) fibers of biceps brachii in 5 MB (C and D) and 4 and posttraining, respectively. The fluctuations in fiber FB (A and B) before and after 24 wk of heavy resistance training. RESISTANCE TRAINING AND MUSCLE 1519

DISCUSSION If Part of the population of small fibers in our current study was the result of fiber proliferation, as i .s the case in Muscle strength, muscle mass, and mean fiber size an avian model of stretch-overload (3, 9, 17), then the have been shown to be greater in highly competitive MB lack of increase in these small fibers would indicate that than in highly competitive FB (7,8), and they are greater fiber proliferation did not occur during the 24 wk of in resistance-trained males (6,8,14,15,18,19) and resis- training. Furthermore, our estimates of total fiber num- tance-trained females (7, 8, 25, 26) than in control sub- ber were not changed after training and therefore sup- jects (22, 27). Nevertheless, the longitudinal effects of port this conclusion. Nevertheless, we recognize that our resistance training on muscle characteristics have not techniques may not have been sensitive enough to detect been studied extensively in females. To our knowledge, small changes in fiber number or new fiber formation. In the current study is the first investigation of longitudinal support of this, Sjijstriim and colleagues (24) have re- effects of heavy resistance training on muscle fiber char- cently suggested that while fiber proliferation probably acteristics of highly competitive MB or FB. occurs in human muscle, detection of new fibers from Selective type II hypertrophy has been reported in biopsy samples will be difficult because of the low fre- some but not all muscles of males (4-7,18,22,27). Selec- quency of this process. In addition, our techniques for tive type II hypertrophy has also been found after resis- estimating total fiber number from CT scans and needle tance training in the quadriceps muscles of women (25, biopsies might not be precise enough because the variabil- 26). Nevertheless, our data indicate that FB did not ob- ity in estimates was large. tain a selective type II hypertrophy by the 5th yr of train- A small but significant increase in flexor (brachialis + ing that preceded this study, nor was the relationship biceps brachii) muscle CSA (-8%) was found after 24 wk between type I and type II fibers altered by the training of training. Such small improvements in muscle mass in this study. Type II-to-type I fiber area ratios were simi- may be important to the overall muscle appearance of lar in FB and male and female controls, and these were the bodybuilder, so these small changes should not be significantly lower than that in MB before the training ignored. When CSA of biceps and brachialis muscles was program in the current study. Because neither MB or FB determined independently] the relative increase in the altered their fiber characteristics with training, we were brachialis was significant for both groups after 24 wk of unable to establish whether preferential type II hyper- training. It is therefore possible that the exercise proto- trophy was possible in the biceps of women bodybuilders. col stressed the brachialis muscle to a greater extent than Furthermore, we cannot rule out the possibility of a sub- the training program that the bodybuilders had previ- ject sampling error because there were only five FB exam- ously used. Because the brachialis inserts into the coro- ined in this study. noid process of the ulna, it is equally efficient whether F, was greater in MB relative to other groups, but not the wrist is pron ated or supinated. In contrast, the biceps all fibers were hypertrophied relative to FB. In this study are most effecti .ve as an elbow flexo r when the h .and is we confirm our previous observations that MB have a supinated. It is therefore possible that the brachialis re- greater frequency of fibers that are ~2.0 X lo3 pm2 rela- ceived more total work than the biceps brachii during the tive to FB. This information is not readily apparent from exercise program. An alternative explanation is that the F, data alone. Nevertheless, 24 wk of resistance training biceps muscles may have received a greater amount of did not alter the frequency-area relationship in either training before this study and therefore would have been athlete group, and it did not alter the frequency of small closer to their peak potential for muscle mass than the fibers. Although electron-microscopic examination of brachialis. these small fibers has not been completed, light-micro- It is worth noting that the small increases in muscle scopic examination of the tissue has failed to provide any mass noted above did not contribute to an improvement evidence that suggests these are injured, atrophied, or in muscle strength as measured by the cybex isokinetic denervated fibers. This is different from reports by dynamometer. However, isokinetic and isometric tests Staron and colleagues (25), who observed apparently may not have been sensitive or specific enough to detect atrophic or denervated fibers in the quadriceps of women changes in muscle strength, because this type of strength after a period of resistance training. These authors (25) testing differed from the free weight type of training that suggested that the small and atrophic fibers found in the subjects employed in the study. Although the body- their posttraining biopsy were from damage incurred builders increased the resistance of their 10 RM free- during the pretraining biopsy procedure. If part of the weight exercises by lo-15% over the 24 wk of the current small fiber population in our study were the result of study, their exercise biomechanics ( e. g ., elbow or fibers damaged by the previous biopsy, we would have shoulder position or swinging the weight slightly) were expected a greater frequency of small fibers in all subject not rigidly controlled. A better assessment of strength groups during the second biopsy. Although we did not could have been made if we had carefully controlled the find an increase in small fibers or any evidence of damage cadence, arc, and movement pattern of the resistance to the muscle fibers, we did find a large variation in fiber and limb, hand position, and other biomechanical param- size both before and after training in our subjects. This eters during a free-weight one RIM. However, because emphasizes the importance of selecting enough fibers to highly competitive bodybuilders do not train in such a obtain a representative sample of the data (7) and pres- rigid and consistent manner, we did not expect free- enting fiber area as frequency histograms and not just as weight one RM strength to provide additional data per- mean fiber size. taining to the athletes improvement in strength. 1520 RESISTANCE TRAINING AND MUSCLE Nevertheless, because rigidly monitored one RM free- tained a larger number of subjects if we had not chosen weight assessments of strength would closely parallel the such a high caliber of athlete. We fully recognize the dif- free-weight exercise patterns in other athletes, such as ficulties that are imposed on data interpretation from weight lifters or power lifters, carefully controlled free- such few subjects in that very large changes would have weight one RM assessments of strength can be consid- had to occur before statistical significance was achieved. ered as both accurate and valid measures of muscle It is, however, important to recognize that we found very strength for these and many other types of athletes (14, small changes in our subjects. In an attempt to determine 15). Hakkinen and colleagues (14,X) found an increase the statistical power for fiber area and strength mea- in dynamic strength and/or muscle CSA in highly com- sures, we have estimated the number of subjects that petitive weight lifters when they used free-weight assess- would be necessary to obtain statistical significance with ments to indicate an improvement in performance, a = 0.05 and ,& = 0.10. For these calculations we esti- rather than isokinetic or isometric tests. Because success- mated variance in fiber area, muscle CSA, and muscle ful competitive weight lifting is judged on a consistent torque data from other studies using bodybuilders (1, and specific lifting technique, it is likely that free-weight 4-8, 18, 19, 22, 28). Because the standard deviations in one RM assessments of strength are a good indicator for fiber area are greater in bodybuilders than nontrained performance in those subjects. However, the primary subjects because of the existence of small as well as large goal for a bodybuilder of either sex is to achieve maximal fibers in bodybuilders (6-8) and because we found only muscle mass and minimal body fat, and in general one small changes in fiber area pre- to posttraining, 30-50 RM strength is not important. Therefore bodybuilders subjects would have been necessary to obtain statistical do not attempt to perform biomechanically strict train- significance in the current study. Stated differently, if ing programs, as is the case in weight lifting. Rather, 200 fibers had been examined in 20 subjects from each highly competitive bodybuilders will often alter the bio- group, we would have needed a pre- to posttrained type I mechanical movement between training sessions on the fiber change of 1,211.7 and 1,461.l pm2 in FB and MB, same exercises by changing hand, elbow, or arm position respectively, to obtain statistical significance. Similarly, or by altering exercise cadence. It is therefore important if 200 type II fibers had been examined in 20 FB and 20 to make a distinction between bodybuilders and weight MB, we would have required pre- to postchange of 978.2 lifters and, furthermore, it is not appropriate to combine and 2,369.4 pm2 in FB and MB, respectively, to achieve bodybuilders and weight lifters or power lifters into a statistical significance. Although less dramatic, 20 sub- single group for analyses of longitudinal or cross-sec- jects in each group would have detected a change in bi- tional training effects. Thus it is not clear that carefully controlled one RM strength assessments necessarily ceps CSA of 2.9 cm2 in MB and 1.5 cm2 in FB. Twenty would have provided us with different results than isoki- subjects in each group would have detected an isometric netic and isometric testing in the bodybuilders of this or isokinetic torque change of 22.5 N l m in MB and 18.9 study. Nevertheless, it is possible that isokinetic N l m in FB groups. Because the actual changes in fiber and/or isometric testing u .nderestimated training-in- areas, muscle CSA, and strength over 24 wk of training duced changes in strength of the highly competitive were far less than these estimated changes that would be bodybuilders. If this was the case, specific tension needed if 20 subjects from each gender had been exam- (torque/cm2 of muscle) may have increased in MB and ined, it is likely that adding more subjects to our study FB as a result of better recruitment of motor units (21), would not have produced a different result. Nevertheless, because there was no evidence for increases in fiber or we cannot discount the possibility that if the power of the muscle size. statistical tests of significance had been higher, the out- The training program had little effect on fiber size or come might have been different. fiber number in either MB or FB. Thus it seems that In summary, highly competitive MB and FB showed a after reaching a very high level of competition, these lack of improvement in fiber size, fiber number, or mus- bodybuilders are only maintaining their existing muscle cle CSA over 24 wk of resistance training. This suggests mass rather than adding to it. This observation is consis- that after reaching a highly competitive status, further tent with findings of highly competitive weight lifters, improvements are minimal. Nevertheless, although where muscle characteristics were relatively static when small improvements in muscle structure and mass may these athletes were examined over intervals of 1 and 2 yr be statistically insignificant, they may represent a physio- (14, 15). logically important factor in determining success in Although three males and two females used steroi .ds highly competitive levels of bodybuilding competition for during this study, there were large overlaps in th .e data of both genders. steroid vs. nonsteroid user in fiber number, fiber CSA, muscle CSA, and the volume density of noncontractile We appreciate the cooperation of the bodybuilders who participated tissue (i.e., steroid users did not consistently have the in the study. We also thank Patricia Gilford for expertise and technical largest or smallest data points). However, the use of ana- help in obtaining computerized topographic scans. bolic steroids did not further alter the muscle character- Present addresses: W. H. Grumbt, Physician Assistant Program, Trevecca Nazarene College, PO Box 1741, Nashville, TN 37210; J. istics of these athletes over 24 wk of training. Stray-Gundersen, Baylor/UT Southwestern Sports Science Research The low number of subjects examined in this study Center, 411 N. Washington, Suite 3000, Dallas, TX 75246. resu lted in part l from an attempt to obtai n bodybuilders Address for reprint requests: S. E. Alway, Exercise Physiology and that were state champions or better . We could have ob- Neuromuscular Laboratories, School of Health, Physical Education, RESISTANCE TRAINING AND MUSCLE 1521

and Recreation, The Ohio State University, 337 W. 17th Ave., Colum- strength training in two years. J. Appl. Physiol. 65: 2406-2412, bus, OH 43210. 1988. 16. HOLLAND, G. J. A review of in vivo quantitative analysis of body Received 22 May 1991; accepted in final form 25 October 1991. composition. In: AAHPERD Research Consortium Symposium Papers, edited by M. Adrian and J. Brane. Washington, DC: AAH- PERD, 1980, p. 46-75. REFERENCES 17. KENNEDY, J. M., B. R. EISENBERG, S. K. REID, L. J. SWEENEY, AND R, ZAK. Nascent muscle fiber appearance in overloaded 1. ALWAY, S. E. Is fiber mitochondria volume density a good indicator chicken slow-tonic muscle. Am. J. Anat. 18: 203-215, 1988. of muscle fatigability to isometric exercise? J. Appl. Physiol. 70: 18. MACDOUGALL, J. D., D. G. SALE, S. E. ALWAY, AND J. R. SUTTON. 2111-2119,199l. Muscle fiber number in biceps brachii in bodybuilders and control 2. ALWAY, S. E. Perpetuation of muscle fibers after removal of subjects. J. Appl. Physiol. 57: 1399-1403, 1984. stretch overload in the Japanese quail. Am. J. Physiol. 260 (CeZZ 19. MACDOUGALL, J. D., D. G. SALE, G. C. B. ELDER, AND J. R. SUT- Physiol. 29): C400-C408, 1991. TON. Muscle and ultrastructural characteristics of highly competi- 3. ALWAY, S. E., M. E. DAVIS, AND W. J. GONYEA. Muscle fiber for- tive powerlifters and bodybuilders. Eur. J. Appl. Physiol. Occup. mation and fiber hypertrophy during the onset of stretch-overload. Physiol. 48: 117-126, 1982. Am. J. Physiol. 259 (Cell Physiol. 28): C92-C102, 1990. 20. MORITANI T., AND H. A. DEVRIES. Neural factors versus hyper- 4. ALWAY, S. E., J. D. MACDOUGALL, AND D. G. SALE. Contractile trophy in the time course of muscle strength gain. Am. J. Phys. adaptations after isometric exercise in the human triceps surae. J. Med. Rehabil. 58: 115-130, 1979. Appl. Physiol. 66: 2725-2732, 1989. 21. SALE, D. G. Influence of exercise and training on motor unit activa- 5. ALWAY, S. E., J. D. MACDOUGALL, AND D. G. SALE. Twitch con- tion. Exercise Sport Sci. Rev. 15: 95-151, 1984. tractile adaptations are not dependent on the intensity of isometric 22. SALE, D. G., J. D. MACDOUGALL, S. E. ALWAY, AND J. R. SUTTON. exercise in the human triceps surae. Eur. J. Appl. Physiol. Occup. Voluntary strength and muscle characteristics in untrained men Physiol. 60: 346-352,199O. and women and male,bodybuilders. J. Appl. Physiol. 62: 1786-1793, 6. ALWAY, S. E., J. D. MACDOUGALL, D. G. SALE, J. R. SUTTON, AND 1987. A. J. MCCOMAS. Functional and structural adaptations in skeletal 23. SIRI, W. E. Body composition for fluid spaces and density: analysis muscle of trained athletes. J. Appl. Physiol. 64: 1114-1120, 1988. of methods. In: Techniques for Measuring Body Composition, edited 7. ALWAY S. E., J. STRAY-GUNDERSEN, W. H. GRUMBT, AND W. J. by J. Brozek and A. Henschel. Washington, DC: Natl. Acad. Sci., GONYEA. Contrasts in muscle and myofibers of highly competitive 1961, p. 223-244. male and female bodybuilders. J. Appl. Physiol. 67: 27-311, 1989. 24. SJOSTROM, M., J. LEXELL, A. ERIKSSON, AND C. C. TAYLOR. Evi- 8. ALWAY, S. E., J. STRAY-GIJNDERSEN, W. H. GRUMBT, AND W. J. dence of fibre hyperplasia in human skeletal muscles from healthy GONYEA. Muscle cross-sectional area and torque in resistance- young men? Eur. J. Appl. Physiol. Occup. Physiol. 62: 301-304, trained subjects. Eur. J. Appl. Physiol. Occup. Physiol. 60: 86-90, 1991. 1990. 25. STARON, R. S., M. J. LEONARDI, D. L. KAFUIPONDO, E. S. MALICKY, 9. ALWAY, S. E., P. K. WINCHESTER, M. E. DAVIS, W. J. GONYEA. J. E. FALKEL, F. C. HAGERMAN, AND R. S. HIKIDA. Strength and Regionalized adaptations and fiber proliferation in stretch-induced skeletal muscle adaptations in heavy resistance-trained women muscle enlargement. J. Appl. Physiol. 66: 771-781, 1989. after detraining and retraining. J. Appl. Physiol. 70: 631-640,199l. 10. BEDELL, G. N., R. MARSHALL, A. B. DEBOIS, AND J. H. HARRIS. 26. STARON, R. S., E. S. MALICKY, M. J. LEONARDI, J. E. FALKEL, F. C. Measurement of the volume of gas in the gastrointestinal tract. HAGERMAN, AND G. A. DUDLEY. Muscle hypertrophy and fast fiber Values in normal subjects and ambulatory patients. J. Clin. Inuest. type conversions in heavy resistance-trained women. Eur. J. Appl. 35: 336-345,1956. Physiol. Occup. Physiol. 60: 71-79, 1990. 27. TESCH, P. A. Skeletal muscle adaptations consequent to long-term 11. BROWN, C. H., AND J. H. WILMORE. The effects of maximal resistance training on the strength and body composition of women heavy resistance exercise. Med. Sci. Sports Exercise 20, Suppl.: S132-S134,1988. athletes. Med. Sci. Sports Exercise 8: 174-177, 1974. 28. TESCH, P. A., AND L. LARSSON. Muscle hypertrophy in body- 12. BUSKIRK, E. R. Underwater weighing and body density, a review of builders. Eur. J. Appl. Physiol. Occup. Physiol. 49: 301-306, 1982. procedures. In: Techniques for Measuring Body Composition, edited 29. WEIBLE, E. R. Stereological Methods. London: Academic, vol. 1, by J. Brozek and A. Henschel. Washington, DC: Natl. Acad. Sci., 1979. 1961, p. 90-106. 30. WILMORE, J. H. Alterations in strength, body composition and 13. FROESE, E. A., AND M. E. HOUSTON. Torque-velocity characteris- anthropometric measures consequent to a ten-week weight train- tics and muscle fiber type in human vastus lateralis. J. Appl. Phys- ing program. Med. Sci. Sports Exercise 6: 133-138,1974. iol. 59: 309-314, 1985. 31. WILMORE, J. H., P. A. VODAK, R. B. PARR, R. N. GIFUNDOLA, AND 14. H~;KKINEN, K., P. V. KOMI, M. ALI?N, AND H. KAUHANEN. EMG, J. E. BILLING. Further simplification of a method for determina- muscle fibre and force production characteristics during a 1 year tion of residual lung volume. Med. Sci. Sports Exercise 12: 216-218, training period in highly competitive weightlifters. Eur. J. Appl. 1980. Physiol. Occup. Physiol. 56: 419-427, 1987. 32. WINTER D. A., R. P. WELLS, AND G. W. ORR. Errors in the use of 15. H~~KKINEN, K., A. PAKARINEN, M. ALAN, H. KAUHANEN, AND P. V. isokinetic dynamometers. Eur. J. Appl. Physiol. Occup. Physiol. 46: KOMI. Neuromuscular and hormonal adaptations in athletes to 397-408,198l.