Eur J Appl Physiol (2006) 98:546555 DOI 10.1007/s00421-006-0300-zORIGINAL ARTICLE

Hypertrophy with unilateral resistance exercise occurs without increases in endogenous anabolic hormone concentration

Sarah B. Wilkinson Mark A. Tarnopolsky Emily J. Grant Caroline E. Correia Stuart M. Phillips

Accepted: 28 August 2006 / Published online: 14 September 2006 Springer-Verlag 2006

Abstract We aimed to gain insight into the role thatthe transitory increases in anabolic hormones play inmuscle hypertrophy with unilateral resistance training.Ten healthy young male subjects (21.8 0.4 years,1.78 0.04 m, 75.6 2.9 kg; mean SE) engaged inunilateral resistance training for 8 week (3 days/week).Exercises were knee extension and leg press performedat 8090% of the subjects single repetition maximum(1RM). Blood samples were collected in the acuteperiod before and after the Wrst training bout and fol-lowing the last training bout and analyzed for total tes-tosterone, free-testosterone, luteinizing hormone, sexhormone binding globulin, growth hormone, cortisol,and insulin-like growth factor-1. Thigh muscle crosssectional area (CSA) and muscle Wbre CSA by biopsy(vastus lateralis) were measured pre- and post-training.Acutely, no changes in systemic hormone concentra-tions were observed in the 90 min period followingexercise and there was no inXuence of training on theseresults. Training-induced increases were observed intype IIx and IIa muscle Wbre CSA of 22 3 and13 2% (both P < 0.001). No changes were observedin Wbre CSA in the untrained leg (all P > 0.5). Whole

muscle CSA increased by 5.4 0.9% in the trained leg(P < 0.001) and remained unchanged in the untrainedleg (P = 0.76). Isotonic 1RM increased in the trainedleg for leg press and for knee extension (P < 0.001). Nochanges were seen in the untrained leg. In conclusion,unilateral training induced local muscle hypertrophyonly in the exercised limb, which occurred in theabsence of changes in systemic hormones that ostensi-bly play a role in muscle hypertrophy.

Keywords Testosterone Growth hormone Strength Weightlifting

Introduction

Muscle hypertrophy resulting from resistance trainingarises due to an accumulation of multiple post-exerciseperiods of positive protein balance (Phillips 2004; Phil-lips et al. 2005; Rennie et al. 2004; Rennie and Tipton2000). Evidence showing that provision of exogenousanabolic hormones to humans, such as testosterone (invarious forms) and growth hormone, is very stronginsofar as testosterones action as an anabolic agent isconcerned (Bhasin et al. 1996; Sinha-Hikim et al. 2002,2003b). What is also clear is that when supraphysiolog-ical doses of testosterone are administered that theysynergistically interact with resistance exercise to pro-mote lean mass and strength gains (Bhasin et al. 1996).Whether the same can be said of the rise in endoge-nous hormones is not clear. Versus testosterone, how-ever, it is entirely equivocal as to whether growthhormone is a hypertrophy-promoting hormone (Langeet al. 2002; Rennie 2003). Various studies have also beenconducted examining the endogenous concentration

S. B. Wilkinson E. J. Grant C. E. Correia S. M. Phillips (&)Exercise Metabolism Research Group, Department of Kinesiology, McMaster University, 1280 Main St. West, Hamilton, ON, Canada, L8S 4K1e-mail: phillis@mcmaster.caURL: http://www.mcmaster.ca/kinesiology/

M. A. TarnopolskyPediatrics and Neurology, McMaster University, Hamilton, ON, Canada123

DamianoEvidenzia

DamianoSottolinea

DamianoTesto digitatonella misura in cui

DamianoEvidenzia

Eur J Appl Physiol (2006) 98:546555 547changes in systemic hormones that are proposed tohave an active role in skeletal muscle hypertrophy andpotentially also in strength gains (Ahtiainen et al. 2003,2005; Hakkinen et al. 1998, 2001; Kraemer et al. 1998,2001; McCall et al. 1999; Nindl et al. 2001; Raastadet al. 2001, 2003). Ultimately, however, the causativerole that acute and/or chronic changes in endogenousanabolic hormone concentrations, induced as a resultof resistive exercise, play in hypertrophy is diYcult todetermine.

The unilateral model of resistance training inhumans is one that has been used to show eVects ofvarious training protocols and examine strength gainsas well as cross-education eVects, (i.e, strength gains ina contralateral untrained limb) (Hakkinen et al. 1996;Hubal et al. 2005; Kim et al. 2005; McCurdy et al.2005; Munn et al. 2005). The unilateral model is con-venient in that it provides for an untrained contralat-eral control limb for comparison that is a within-subject parameter insofar as statistical analysis in con-cerned. This within-person comparison provides a sig-niWcant advantage from the standpoint of reducingvariability, with the assumption that within-personvariability is less than between person, and ultimatelywould allow more subtle diVerences in muscle andmuscle Wbre morphology as well as molecular eventsto be detected (Tarnopolsky et al. 2006). The unilat-eral model has been shown to result in hypertrophy(Hubal et al. 2005; Kim et al. 2005; Shepstone et al.2005) of the trained arm with no signiWcant change, orcomparatively minor changes, in the contralaterallimb. Using the unilateral design, detecting strengthgains is a more complicated prospect due to the cross-education eVect; see Munn et al. (2004) for a recentmeta analysis of this phenomenon.

We chose that the unilateral model since one caninduce local hypertrophy in the ipsilateral limb andexamine strength changes due the combined action ofneuromuscular and hypertrophic mechanisms. Thecontralateral limb serves as a control not for strengthgains, which are contaminated due to cross-education,but for any hypertrophy due to humoral mechanisms,(i.e., hormonal spillover that might potentially aVectgains in strength). Thus the purpose of this study wasto use unilateral resistance training to examine hyper-trophy in the trained limb along with changes in thecontralateral untrained limb and measure changes inendogenous systemic hormones to determine whetherthese changes play a role in the hypertrophic response,in either the trained our untrained limb. Our workinghypothesis was that unilateral training of the quadri-ceps femoris would not induce a signiWcant change inendogenous hormone concentrations but that hyper-

trophy would occur in the trained limb nonetheless.We also hypothesized that the untrained limb wouldnot exhibit hypertrophy and would show lesserstrength gains compared to the trained limb.

Methods

Subjects

The study was approved by the Research Ethics Boardof Hamilton Health Sciences and McMaster Universityand conformed to the regulations laid out in the decla-ration of Helsinki on the use of human subjects inresearch. All subjects gave their written consent to allprocedures prior to participating. Subjects were tenyoung 21.8 0.4 years (mean SE) males (1.78 0.04 m, 75.6 2.9 kg) who were deemed healthy basedon their response to a general health questionnaire.Participants were all non-smokers and had no previ-ous history of resistance training and participated in aminimal amount of physical activity (no more than2 h/week).

Testing

At least 2 week prior to beginning the training protocolsubjects reported to the Exercise Metabolism ResearchLaboratory for orientation to all procedures. At thistime preliminary measures were made to estimate iso-tonic single repetition maximum (1RM) as well as iso-metric 1RM (Biodex-System 3, Biodex MedicalSystems, Shirley, NY). Subjects were also familiarizedwith the guided motion training equipment for legpress and knee extension (Nautilus, Vancouver, WA).

To estimate isotonic 1RM for the knee extensionsubjects were seated with their hips at 90 and theirback against a back rest that was at a 30 angle fromhorizontal (all tests were performed using a Nautilusguided motion machine; Tulsa, OK). The fulcrum ofthe machine was aligned with the lateral aspect of themidline of the subjects knee. The leg pad of themachine was positioned 5 cm above the subjectsankle and the rotation arm was positioned so that thesubjects knee was bent to 90. A full repetition waswhen the subject was able to move the weight throughan arc of 80 (from 90 to 170). The subjects set-tings for pin placements for seating in the machinewere recorded and kept constant throughout the study.Subjects warmed up using a light weight and per-formed 810 repetitions. Subjects then performeda single best eVort at a weight estimated to be thesubjects 1RM based on body weight and height by an123

DamianoEvidenzia

DamianoEvidenzia

DamianoEvidenzia

DamianoEvidenziato

DamianoEvidenziato

DamianoEvidenziato

DamianoSottolinea

DamianoTesto digitatocomunque

DamianoEvidenzia

548 Eur J Appl Physiol (2006) 98:546555experienced trainer. If the subject could not lift theweight, the weight was lowered and if the subject couldlift the weight then it was raised. Ultimately, a 1RMwas determined in this way. This estimated 1RM waschecked on two subsequent occasions by simply havingthe subject report to the lab and asking them to oncemore perform a 1RM at the previously determinedweight. For a true 1RM to be determined the coeY-cient of variation (CV) between two attempts had to beless than 5%. For no subject did it require more thanthree attempts to determine this 1RM.

Isometric 1RM was determined in two sessionsusing the Biodex dynamometer. Subjects were seatedin the dynamometer with their hips at 90 and theirshoulders strapped with harness straps that preventedany upper-body movement. The fulcrum of the rota-tion arm was positioned to align with the midline of thesubjects knee. The subjects knee was bent to an angleof 120 and the pad of the rotation arm was positioned5 cm above the subjects ankle. Isometric 1RM wasdetermined by having the subject maintain a maximaleVort for knee extension and hold that eVort 5 s afterwhich subjects were allowed to rest for 2 min beforeattempting a second and third 1RM. The single high-est peak value of any one of the three eVorts was takenas the isometric 1RM. Subjects reported back to thelaboratory for repeat testing using the same proce-dures 34 days after the Wrst 1RM estimate and if theCV for 1RM determined in these two sessions wasgreater than 5% the procedure was repeated a thirdtime, which was necessary for only one subject.

Leg press 1RM (Nautilus, Tulsa, OK) was deter-mined in a seated position with the subject positionedso that their back rested against the back rest at 30from horizontal and the angle between their torso andtheir legs (at their hips) was 70. Subjects had theirknee bent at an angle of 90 and their foot placed on afoot plate so that a full 1RM resulted in full leg exten-sion to 180. The same procedure to determine 1RM asfor the knee extension was followed for the leg press.A true 1RM was determined for each subject withinthree visits to the lab, or less.

Training

Subjects had their legs randomly assigned to be trainedin a counterbalanced manner so that Wve of ten sub-jects trained their dominant leg and the Wve their non-dominant leg, based on dominance from 1RM strength.Training was conducted 3 days/week (Monday,Wednesday, and Friday) for a total of 8 week. Eachsubject completed all training sessions. The intensity ofthe training was set at 8090% of 1RM, which meant

that subjects usually reached voluntary fatigue within610 repetitions per set. Subjects recovered for 3 minbetween sets. In the Wrst 4 weeks of the training sub-jects complete 3 sets per training session and in weeks58 they completed 3 sets of 810 repetitions and oneset to voluntary fatigue (usually 510 repetitions).Training was progressive in nature with subjects vol-untary isotonic 1RM being retested every 2 weeks inthe trained leg. Intensity was adjusted to ensure thatsubjects were always lifting loads between 80 and 90%of 1RM. The untrained leg was tested in the same man-ner for 1RM throughout the protocol.

Biopsies

Percutaneous muscle biopsies were obtained underlocal anesthesia (2% xylocaine) using a 5 mm Berg-strom biopsy needle custom modiWed for manual suc-tion as detailed previously (Kim et al. 2005; Shepstoneet al. 2005; Stewart et al. 2004; Yasuda et al. 2005).Muscle was mounted in optimal cutting temperature(OCT) medium (Sakura Finetechnical, Osaka, Japan)with Wbres oriented in the plane perpendicular towhich it was to be cut. Detailed methods and proce-dures yielding the descriptive variables for Wbre size,type, percentage area, and myosin heavy chain compo-sition have been described in detail previously else-where (Kim et al. 2005; Shepstone et al. 2005; Stewartet al. 2004; Yasuda et al. 2005).

Muscle computerized tomographic scans

Subjects had both their trained and untrained legsscanned both pre- and post-training. Pre-training thescans were performed at least 2 days after any exerciseor testing procedures and the scans were performed onboth legs simultaneously. A General Electric CTIScanner (GE, Milwaukee, WI) was used to performcomputerized tomographic (CT) scans at baseline andafter cessation of training. A scout scan was taken ofthe lower limbs to determine the femur and tibialengths, and 5 mm slices were taken at two sites: 60%of femur length, starting from the distal end and mea-suring proximally. The system parameters used were asfollows: slice thickness 5 mm, pixel matrix 512 512,and exposure factors of 120 kV, 200 mA and standardreconstruction algorithm. CT scans were analyzedusing a validated software program (BonAlyse 1.3,BonAlyse Oy, Jyvaskyla, Finland), according to themanufacturers instructions. Thresholds 270 to 101HounsWeld Units (HU) were used to identify fat, andthresholds 101 to 270 HU were used to identifymuscle. BonAlyse was used to calculate muscle cross123

DamianoEvidenzia

DamianoEvidenzia

DamianoEvidenzia

Eur J Appl Physiol (2006) 98:546555 549sectional area (CSA) (mm2). We have determined thatmuscle CSA obtained from CT scans using our scannercan be measured with a CV of less than 1% in repeatscans with the same person across several weeks.

Hormone measurements

Prior to the Wrst training workout and the last trainingbout subjects had a venous catheter inserted into a dor-sal hand vein. The hand was kept warm using a heatingblanket to maintain an arterialized sample of blood onwhich to measure hormone concentrations. Subjects hadblood samples drawn immediately prior to performingthe training bout (pre), immediately after the exercisebout (post), and then again at 30, 60, 90, and 120 minpost-exercise. Blood was collected in heparinized tubesas well as additive-free tubes. All tubes were kept on icebrieXy (never more than 5 min for plasma and 10 minfor serum) before being spun in a refrigerated centrifuge(4C, 4,500 rpm) for 10 min to isolate plasma (heparin-ized tubes) or serum (additive-free tubes). Plasma andserum was stored at 80C prior to analysis.

Plasma total testosterone (T), free testosterone (Tf),luteinizing hormone (LH), sex-hormone binding glob-ulin (SHBG), growth hormone (GH), and cortisol (C)were analyzed using commercially available radioim-munoassay kits (Diagnostics Products Corporation,Los Angeles, CA) and manufacturers instructions wereclosely followed. To minimize any intra-assay CV allpre- and post-training samples from each subject wereanalyzed in duplicate on the same day. A between sam-ple CV (on pairs) of less than 5% was taken as goodagreement and the values were averaged to yield a sin-gle concentration. On pooled plasma within-sampleCVs ranged from as low as 0.3% to no higher than4.8% on 68 replicates.

IGF-1 concentrations were measured in duplicate byimmunoassay (Quantikine IGF-1 Immunoassay, R&DSystems, Minneapolis, MN); recombinant human IGF-1 was used to generate the standard curve. The mean

intra-assay CV, determined by assaying the IGF-1 con-centration in a number of samples analyzed as repli-cates was less than 3.0%. The mean minimal detectableconcentration of IGF-1 in this assay is 0.026 ng/ml.

Blood metabolite

Whole blood lactate was determined by adding hepa-rinized plasma to 0.6M perchloric acid to deproteinizethe sample. The sample was then spun (15 min,4,500 rpm, 4C) to pellet the proteins and to the super-natant, which was kept on ice, was added 1.25MKHCO3 to neutralize the sample. Again the resultingsalt was spun (15 min, 4,500 rpm, 4C) down and theresulting supernatant was collected and stored at80C until analysis. Analysis of whole blood lactateconcentration was performed as described previously(Phillips et al. 1995).

Statistics

Data were analyzed using a two-way repeated mea-sures analysis of variance with time on multiple levelsas well as training status (trained vs untrained) of theleg as factors. SigniWcant F ratios were further exam-ined using Tukeys posthoc test to isolate the diVer-ences. SigniWcance was set at P < 0.05. Data arepresented and means SE.

Results

Strength

Isotonic 1RM increased as a result of training for theleg press and for knee extension (Table 1; bothP < 0.001). In addition, a moderate cross-educationeVect was also observed for knee extension 1RM suchthat the contralateral untrained limb showed a15 3% increase (P < 0.05; Table 1).

Table 1 Maximal isotonic strength for both the leg press and knee extension exercises in both the control and trained leg

Values are means SE (N = 10). 1RM single repetition maximum, kg kilogram. Means with diVerent letters are signiWcantly diVerent*SigniWcantly diVerent from control at the same time point (P < 0.01), 9 P < 0.001. Means with diVerent letters are signiWcantly diVerentfrom each other (P < 0.05)

Pre 3 5 7 Post

Leg press 1RM (kg)

Control 143 21a 142 26a 145 26a 143 27a 147 25aTrained 141 26a 149 26b* 159 30c* 161 27c9 166 25d9

Knee extension 1RM (kg)Control 47 8a 50 9a 51 9a 52 9b 54 8bTrained 47 7a 52 7ab 59 8b* 63 11c9 68 12d9123

DamianoEvidenzia

550 Eur J Appl Physiol (2006) 98:546555Isometric 1RM increased by 12 2% in the trainedleg (PRE = 256 46 N m, POST = 286 29 N m; P /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 150 /GrayImageDepth -1 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 600 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputCondition () /PDFXRegistryName (http://www.color.org?) /PDFXTrapped /False

/Description >>> setdistillerparams> setpagedevice