GH Administration Changes Myosin Heavy Chain Isoforms in Skeletal Muscle But Does Not Augment Muscle Strength or Hypertrophy, Either Alone or Combined with Resistance Exercise Training in Healthy Elderly Men
Kai Henrik Wiborg Lange, Jesper Løvind Andersen, Nina Beyer, Fredrik Isaksson, Benny Larsson, Michael Højby Rasmussen, Anders Juul, Jens Bülow and Michael Kjær
Sports Medicine Research Unit (K.H.W.L., N.B., F.I., M.K.), Bispebjerg Hospital, DK-2400 Copenhagen NV, Denmark; Copenhagen Muscle Research Centre (K.H.W.L., J.L.A.), Department of Molecular Muscle Biology, Rigshospitalet, DK-2100 Copenhagen Ø, Denmark; Team Danmark Test Center (B.L.), Bispebjerg Hospital, DK-2400 Copenhagen NV, Denmark; Clinical Drug Development (M.H.R.), Novo Nordisk, DK-2880 Bagsvaerd, Denmark; Department of Growth and Reproduction (A.J.), Rigshospitalet, DK-2100 Copenhagen Ø, Denmark; and Department of Clinical Physiology (J.B.), Bispebjerg Hospital, DK-2400, Copenhagen NV, Denmark
Address all correspondence and requests for reprints to: Kai H. W. Lange, M.D., Sports Medicine Research Unit, Building 8, Bispebjerg Hospital, Bispebjerg Bakke 23, DK-2400 Copenhagen NV, Denmark. E-mail:
[email protected].
Abstract
GH administration, either alone or combined with resistance exercise training (RT), has attracted interest as a means of increasing muscle mass and strength in the elderly. In the present study, 31 healthy, elderly men [age, 74 ± 1 yr (mean ± SEM)] were assigned to either RT [3 sessions/wk, 3–5 sets of 8–12 repetition maximum (RM)/session] + placebo (n = 8), RT + GH (n = 8), GH (n = 8), or placebo (n = 7) in a randomized, placebo-controlled, double-blinded (RT + placebo and RT + GH) or single-blinded (GH or placebo) design. Measurements of: 1) isokinetic quadriceps muscle strength; 2) quadriceps muscle power; 3) quadriceps muscle fiber type, size, and myosin heavy chain (MHC) composition; 4) quadriceps cross-sectional area (CSA) [nuclear magnetic resonance imaging (NMRI)]; 5) body composition (dual-energy x-ray absorptiometry scanning); and 6) GH-related serum markers were performed at baseline and after 12 wk. The final GH dose was 1.77 ± 0.18 IU·d-1 (7.2 ± 0.8 µg·kg-1·d-1). GH alone had no effect on isokinetic quadriceps muscle strength, power, CSA, or fiber size. However, a substantial increase in MHC 2X isoform was observed with GH administration alone, and this may be regarded as a change into a more youthful MHC composition, possibly induced by the rejuvenating of systemic IGF-I levels. RT + placebo caused substantial increases in quadriceps isokinetic strength, power, and CSA; but these RT induced improvements were not further augmented by additional GH administration. In the RT + GH group, there was a significant decrease in MHC 1 and 2X isoforms, whereas MHC 2A increased. RT, therefore, seems to overrule the changes in MHC composition induced by GH administration alone. Changes in body composition confirmed previous reports of decreased fat mass, increased fat-free mass, and unchanged bone mineral content with GH administration. A high incidence of side effects was reported. Our results do not support a role for GH as a means of increasing muscle strength or mass, either alone or combined with RT, in healthy elderly men; although GH administration alone may induce changes in MHC composition.
GH IS AN anabolic hormone capable of increasing muscle mass ( 1). This has been demonstrated clearly in animals, using different models combining GH administration, hypophysectomi, hindlimb suspension, and casting (immobilization) ( 2, 3, 4, 5, 6, 7). In humans, GH administration is known to increase both whole-body and muscle protein synthesis ( 8, 9, 10) and almost unequivocally to increase lean body mass (LBM) and decrease FM ( 11, 12, 13, 14, 15, 16, 17). GH administration has therefore attracted great interest, in the past decade, as a means of increasing LBM and muscle strength, either alone or in combination with different training regimens ( 17, 18, 19, 20, 21). Investigations have predominantly focused on the elderly population, because it is known that the elderly have a decreased function of the GH-axis, and this has been proposed to play a causative role in the decreased muscle mass and strength also observed in the elderly ( 22). Furthermore, if such a causal relationship between GH administration and muscle function exists, the elderly would obviously be the target population that would benefit the most from GH administration.
In patients with GH deficiency, some reports have demonstrated small increases in muscle strength ( 23, 24) and endurance ( 16, 25) with GH therapy. Nevertheless, results have not been encouraging in healthy elderly subjects. Despite undisputed increases in LBM, GH administration alone does not seem to increase aerobic performance [maximal oxygen uptake (VO2max)] or muscle strength in elderly men ( 26) or women ( 27). Furthermore, GH administration, combined with different resistance exercise training (RT) regimens in elderly men, did not augment the strength improvements induced by training alone in two of the three reports published so far ( 17, 19). In one report, however, Welle et al. ( 20) found a significant effect of GH administration on muscle strength when combined with RT in elderly men. Although differences in study design exist, these conflicting results are not immediately intelligible.
Investigations in healthy elderly human subjects, with respect to the effects of GH administration on muscle function, are very limited in number. To our knowledge, the effects of GH administration alone has only been reported in the study by Papadakis et al. ( 26) (n = 26 receiving GH), and only three reports have studied the effects of GH administration combined with RT (n = 10 + 8 + 6 = 24 receiving GH), the latter three giving conflicting results ( 17, 19, 20).
With this background, we conducted a randomized, placebo-controlled, partially double-blinded RT study in elderly men. Furthermore, because no information is available on changes in myosin heavy chain (MHC) composition in this context, we also focused on this issue, because skeletal muscle MHC composition has been shown to correlate with functional muscle characteristics on both single-fiber ( 28) and whole-muscle levels ( 29). Thirty healthy men (>70 yr old) completed a 12-wk supervised RT program, having been assigned to either RT + placebo, RT + GH, GH alone, or placebo alone. We hypothesized that GH administration alone would increase muscle mass and strength and that GH administration combined with RT would increase muscle mass and strength more than RT alone. Furthermore, we hypothesized that GH alone or combined with RT would induce changes in skeletal muscle MHC composition.
Materials and Methods
Subjects
Thirty-one healthy, elderly male subjects were included in the study; age, 74 ± 1 yr (70–82) [mean ± SEM (range)]; height, 174 ± 1 cm (163–186); body weight, 80.8 ± 1.8 kg (63.1–100.3); body mass index, 26.7 ± 0.5 kg·m-2 (20.9–32.8); body fat, 22.9 ± 1.2 kg (10.9–38.9), 28.0 ± 1.0% (14.9–40.0). Before inclusion, each subject underwent a comprehensive medical evaluation, including medical history, physical examination, routine blood tests, and an exercise electrocardiogram. Exclusion criteria were metabolic, cardiac, and malignant disease; anemia; hormonal replacement therapy; and medication with - or ß-blockers.
Ethical approval
Informed consent was obtained according to the Helsinki 2 declaration, and the study protocol was approved by the Ethics Committee for Medical Research in Copenhagen (KF 02-130/97) and by the Danish National Board of Health (journal no. 5312-181-1997).
Experimental protocol
After inclusion, the subjects were randomized, in blocks of four, to either RT or no training and to receive either GH or placebo. Because not all measurements were performed in the group receiving only placebo, the principal investigator was not blinded with respect to the two groups assigned to no training. In this way, GH administration was double-blinded in the two training groups and single-blinded in the two groups assigned to no training. The subjects in the training groups subsequently underwent a 12-wk supervised strength training program. Measurements were made at baseline and after 5 and 12 wk of training and/or GH/placebo administration. The blinding was maintained until data acquisition was completed for the whole study.
Administration of recombinant human GH
Recombinant human GH or placebo (GH, Norditropin PenSet 24; placebo, Norditropin PenSet 24 Placebo; both from Novo Nordisk, Denmark) was administered sc in the thigh, one time per day. Both drug and injection devices were similar for placebo and GH. After thorough instruction, the subjects were able to perform the injections themselves at home, in the evening, before bedtime. Syringes and injection technique were reviewed weekly by a trained technician to ensure compliance. At the commencement of the study, the dose was increased, over 3 wk, to avoid side effects. During the first week, the dose was 0.5 IU·m-2; during the second week, 1.0 IU·m-2; and for the remaining of the period, 1.5 IU·m-2 (12 µg·kg-1·d-1). The subjects were weighed, checked for edema, and questioned about side effects every week by the same physician. If side effects appeared, the dose was reduced by 50%, until the side effects had disappeared or were tolerable to both the subject and the investigators.
Determination of peak oxygen uptake (VO2 peak)
VO2 peak was determined, at baseline, on an electromagnetically braked bicycle ergometer (Technogym, Bikerace, Gambettola, Italy). A protocol starting at 40 W and increasing with 20 W every 2 min until exhaustion was used. Respiratory variables were measured using an AMIS 2001 automated metabolic cart (INNOVISION, Odense, Denmark) and were averaged for each 15-sec period. The mean of the two highest 15-sec values was recorded as VO2 peak.
Dual-energy x-ray absorptiometry (DEXA)
Body composition was determined by DEXA scanning at baseline and after 12 wk. The subjects were scanned, in the morning, after an overnight fast and after having emptied the urine bladder. A DPX-IQ scanner (Lunar Corp., Madison, WI; software version 4.6 C) was used, and the scans were performed at baseline and after 12 wk of training/GH administration. A medium scanning procedure (25 min) was chosen, and the same two trained technicians performed all scans. The same investigator analyzed all scans, using the extended research analysis software provided by Lunar Corp. In the analysis, total body scan was divided into three regions: arms, legs, and trunk. Total and regional scans were further divided into three compartments: fat free mass (FFM), FM, and bone mineral content (BMC).
Muscle biopsies
Muscle biopsies were obtained from the right vastus lateralis muscle, at midthigh level, at baseline and after 12 wk. Sampling was performed in the morning, after an overnight fast; and, in the two training groups, the 12-wk biopsies were obtained 24 h after completion of the last training session. Sampling at 12 wk was performed, 1 cm proximal to baseline sampling. The overlying skin was anesthetized with 1% lidocaine, and sampling was done through an incision using a 5-mm Bergström needle ( 30). A suction device, in conjunction with the biopsy needle, was used to create a negative pressure while sampling, which allowed for a larger sample specimen. Muscle samples were immediately mounted in tissue-tek, frozen in isopentane cooled with liquid N2, and transferred to vials for storage at -80 C until analysis. Biopsies were not obtained from the group receiving only placebo.
Leg extensor power
Measurement of leg extensor power was performed at baseline and after 12 wk in a Nottingham Power Rig, as reported previously ( 31). The subject was in a seated position, and single explosive efforts of the leg extensors accelerated a flywheel from rest. The final speed of the flywheel was used to calculate average power. Measurements were repeated, with a minimum of 10 trials or until no further improvements were observed. Each leg was tested separately, and verbal encouragement and visual feed-back were given. The average of the 3 best trials of the right leg was used as a measure of the maximal leg extensor power.
Isokinetic force (torque)
Knee strength of the strongest leg was measured in a CYBEX 6000 device (Lumex Inc., Ronkonkoma, NY) at baseline and after 12 wk. The subjects sat leaning against a backrest reclined 15° from vertical. They were stabilized, i.e. strapped, at the shoulder, waist, and distal portion of the thigh; and the rotational axis of the dynamometer was aligned with the lateral femoral epicondyle. The lower leg was attached to the load cell positioned proximal to the ankle. Maximal isokinetic concentric knee extension and flexion were measured at an angular velocity of 60°·sec-1, followed by an angular velocity of 180°·sec-1. The subjects were familiarized to the procedure by four warm-up trials, followed by six maximal trials at each velocity, to ensure that peak torque, which was used as a measure of the maximal force, occurred within the trial ( 32). Strong verbal reinforcement was used to achieve an optimal effort level. Data were always corrected for the effect of gravity on the shank, foot, and ankle pad.
NMRI
NMRI was performed at baseline and after 12 wk. 2D T1-weighted fast field echo (TR/TE, 500/14 ms; FOV180; matrix 512 x 512; slice thickness, 6 mm) MR images (Philips, Gyroscan ACS-NT 1.5 T, Best, Holland) were obtained at a level positioned two thirds proximal along an axis connecting a fix point at the tibial eminentia intercondylaris with a fix point at the femoral trochanter major. The left leg was always examined to avoid possible interference with hematomas originating from muscle biopsies. The rectus femoris, vastus lateralis, vastus intermedius, and vastus medialis muscles were identified; and their circumference and cross-sectional area (CSA) were determined. Total quadriceps CSA was calculated by adding the four individual areas together. An estimate of sc fat CSA was obtained by subtracting leg CSA within the muscle fascia from total leg CSA. NMRI was not performed in the group receiving only placebo.
Training program
The subjects in the two training groups performed three different lower-body exercises: leg press, seated knee extension, and seated knee flexion. Exercises were performed as 3–5 sets of 8–12 repetitions. From wk 6–12, the load equaled 8 RM (repetition maximum). Four different upper-body exercises were also performed: pull-down, chest press, abdominal crunch, and back extension. The upper-body exercises were performed as 1 set of 8–15 repetitions. Exercises, except back extensions and abdominal crunches, were performed in weight-loaded strength-training machines (Technogym, Gambettola, Italy). Training was performed 3 times per week, and the total duration of one training session was approximately 30 min. An investigator supervised each training session; 1 RM was determined at baseline and after 12 wk.
Blood sampling
Whole blood was sampled at baseline and after 12 wk, from a cubital vein, into sealed vials without any additives; allowed to clot for 15 min at room temperature; and centrifuged for 15 min at 4 C. The resulting serum was transferred to appropriate tubes and stored at -80 C until analysis for GH-related serum parameters. Sampling was performed in the morning, after an overnight fast, and 24 h after the last training session for the 12-wk sample.
Analytical methods
Biopsies Histochemistry analysis.
Serial sections (10-µm) of the muscle biopsy samples were cut in a cryostat (-20 C), and routine ATPase histochemistry analysis was performed after preincubation at pH 4.37, 4.60, and 10.30 ( 33). Five different fiber types were defined (1, 1/2a, 2a, 2ax, and 2x) according to Staron et al. ( 34, 35), with the modification that fibers termed by Staron et al. as 1c and 2ac (2c) were pooled into one group termed: 1/2a. Cross-sections from baseline and 12-wk biopsies from the same subject were transferred to the same slide and processed for ATPase histochemistry simultaneously. Fibers determined as type 2 fibers, but showing an intermediate staining with pH 4.60 preincubation, were categorized as type 2ab fibers ( 34, 35). These type 2ab fibers covered a wide range, from fibers with only a light staining (i.e. fibers with predominately MHC IIA content) to fibers with a darker staining (i.e. fibers with predominantly MHC IIX content) ( 36); 246 ± 8 fibers were examined in each of the biopsies. Only truly horizontally cut fibers were used in the determination of fiber size. Thus, a restricted number of fibers (170 ± 8 fibers in each biopsy) were used for this analysis ( 36).
Analysis of serial cryosections.
The serial sections were visualized and analyzed using an Olympus Corp. BX40 microscope (Olympus Corp. Optical Co., Ltd, Tokyo, Japan), a Sanyo Hi-resolution Color CCD camera (Sanyo Co., Ltd., Osaka, Japan), and an 8-bit Matrox Meteor Framegrabber (Matrox Electronic Systems Ltd., Québec, Canada), combined with image-analyzing software (Tema, Scanbeam, Hadsund, Denmark) ( 36).
MHC analysis.
MHC analysis was performed on the muscle biopsies, using SDS-PAGE; 10–20 serial cross-sections (20 µm) were cut from each biopsy. Cross-sections for MHC analysis were cut from the biopsies and were placed in 100–200 µl lysing buffer and heated for 3 min at 90 C ( 37); 5–20 µl of the myosin-containing samples were loaded on a SDS-PAGE gel containing 6% polyacrylamide and 30% glycerol. Gels were run at 70 V for 42 h in 4 C. Subsequently, the gels were Coomassie-stained, and MHC isoform content was determined with a densitometric system (Cream 1-D, Kem-En-Tec Aps, Copenhagen, Denmark) ( 36).
Serum Total IGF-I was determined by RIA as previously described ( 38). Briefly, serum was extracted by acid-ethanol and was cryoprecipitated before analysis, to remove interfering IGF binding proteins (IGFBPs). Inter- and intraassay coefficients of variation were less than 9% and 6%, respectively. Details regarding determination of total IGF-I have been presented previously.
IGF-II was determined by an immunoradiometric assay (Diagnostic Systems Laboratories, Inc., Webster, TX). Briefly, this assay is a noncompetitive assay in which the analyte is sandwiched between two antibodies. Samples were pretreated (1:1000) with acid-ethanol extraction to separate IGF-II from its binding proteins before measurement. Inter- and intraassay coefficients of variation were 6.3–10.4% and 4.2–7.2%, respectively ( 39).
IGFBP-3 was determined by an RIA as previously described ( 40). Reagents for the assay were obtained from Mediagnost GmbH (Tübingen, Germany). The sensitivity was 0.29 µg·liter-1 (defined as 3 SD from the mean of the zero standard). Inter- and intraassay coefficients of variation were 10.7% and 2.4% (at bound-to-free ratios of 0.4–0.5), respectively. Details regarding determination of IGFBP-3 have been presented previously ( 41).
Acid-labile subunit (ALS) was determined by a commercially available ELISA (Diagnostic Systems Laboratories, Inc.) ( 42). Standards ranged from 1.09–70 mg·liter-1. In our hands, interassay coefficients of variation (n = 22) were 20.4% (at 2.8 mg·liter-1) and 12.1% (at 17.6 mg·liter-1), respectively. Intraassay coefficients of variation (n = 20) were 8.6% (at 30.1 mg·liter-1) and 7.4% (at 8.4 mg·liter-1), respectively ( 43).
Statistical analysis Data are presented as means ± SEM. Kruskal-Wallis test was used to detect significant differences in baseline characteristics between groups. Wilcoxon’s paired test was used to detect significant changes within groups. Mann-Whitney unpaired test was used to detect significant changes between groups. P < 0.05 (two-tailed) was considered significant.
Results
Subjects
Thirty of 31 included subjects completed the study. One subject abandoned the study after 8 wk because of intolerable side effects attributable to GH administration (pitting leg edema). Baseline characteristics of the remaining 30 subjects, divided into RT + placebo (n = 8), RT + GH (n = 8), GH (n = 7), and placebo (n = 7), are presented in Table 1. There were no significant differences in baseline characteristics among the groups.
View this table:
[in this window]
[in a new window]
Table 1. Baseline characteristics of the 30 subjects completing the study
GH administration
Of the 15 subjects, who completed the study and received GH, 12 experienced side effects. In 10 of these subjects, a dose reduction was necessary. The expected final dose of 2.60 ± 0.05 IU·d-1 (10.6 ± 0.2 µg·kg-1·d-1) was thus reduced to 1.77 ± 0.18 IU·d-1 (7.2 ± 0.8 µg·kg-1·d-1) in the 2 GH groups. Side effects were mainly attributable to fluid retention (varying degrees of pitting leg edema, n = 10; carpal tunnel syndrome, n = 1; so-called triggerfingers, n = 1; transient atrial fibrillation, n = 1; weight gain, n = 1). Of the 15 subjects who completed the study and received placebo, 4 experienced side effects that necessitated a dose reduction. The expected final placebo dose of 2.60 ± 0.07 IU·d-1 (11.1 ± 0.2 µg· kg-1·d-1) was therefore reduced to 2.47 ± 0.08 IU·d-1 (10.6 ± 0.3 µg·kg-1·d-1) in the 2 placebo groups. Side effects were: mild leg edema (n = 2), headache (n = 1), and weight gain (n = 1). The difference in the proportion of side effects between the GH groups and the placebo groups was statistically significant (P = 0.009, Fisher’s exact test); and the final GH dose was significantly lower in the GH groups, compared with the final placebo dose in the placebo groups (P < 0.0025, Mann-Whitney).
GH-related serum parameters
The groups did not differ, with respect to serum IGF-I, IGF-II, IGFBP-3, or ALS at baseline (Table 2). In the two groups receiving GH, the three serum markers IGF-I, IGFBP-3, and ALS increased significantly from baseline to 12 wk, whereas IGF-II almost tended to increase (RT + GH, P < 0.054; GH, P < 0.078) (Table 2). No changes were observed in the two placebo groups from baseline to 12 wk (Table 2).
View this table:
[in this window]
[in a new window]
Table 2. GH-related serum markers
DEXA
There were no differences in body weight, FM (absolute and relative), FFM, or BMC between groups at baseline (Fig. 1). Body weight did not change in either group from baseline to 12 wk (Fig. 1A). RT + placebo and placebo alone did not cause any changes in body composition. However, RT + GH and GH alone decreased total FM (RT+GH, 3.16 ± 1.00 kg, P < 0.0078; GH, 2.27 ± 0.54 kg, P < 0.0156) and increased FFM (RT+GH, 2.52 ± 0.54 kg, P < 0.0078; GH, 2.46 ± 0.51 kg, P < 0.0156) to a similar extent, whereas BMC was unchanged in both groups (both P > 0.5) (Fig. 1). Regional body composition analysis showed that fat percentage was reduced in all three body compartments (arm, leg, trunk) in the RT + GH group and in two body compartments (leg and trunk) in the GH group (Fig. 2). Similarly, FFM percentage was increased in all three body compartments in the RT + GH group and in two body compartments (leg and trunk) in the group receiving only GH (Fig. 2).
View larger version (22K):
[in this window]
[in a new window]
Figure 1. A, Body weight; B, FM; C, FFM; D, BMC (all at baseline and after 12 wk). Bars and error bars represent mean values and SEM, respectively. *, Significant change within a group. See Results for further details.
View larger version (21K):
[in this window]
[in a new window]
Figure 2. Changes in total and regional FM and FFM as a result of the interventions. Changes calculated in percent from values of FM and FFM (measured in percent). Bars and error bars represent mean values and SEM, respectively. *, Significant change within a group.
NMRI
There were no differences in quadriceps CSA or sc femoral CSA between the groups at baseline (quadriceps CSA, P > 0.69; femoral sc CSA, P > 0.81). Quadriceps CSA increased significantly in the RT + placebo and in the RT + GH groups (RT + placebo, 6.3 ± 2.5%, P < 0.05; RT + GH, 10.4 ± 2.7, P < 0.0156), from baseline to 12 wk, but did not change in the group receiving only GH (P > 0.81) (Fig. 3). There was no difference in the magnitude of the increase in CSA between the two training groups (P < 0.16). Sc femoral CSA decreased significantly from baseline to 12 wk in the group receiving only GH (5.2 ± 1.9%, P < 0.0156), tended to decrease in the RT + GH group (6.3 ± 3.0%, P < 0.0781), and was unchanged in the RT + placebo group (0.6 ± 2.1%, P > 0.98) (Fig. 3).
View larger version (21K):
[in this window]
[in a new window]
Figure 3. CSA determined by NMRI. A, Quadriceps CSA at baseline and after 12 wk; B, change in quadriceps CSA from baseline to 12 wk, in percent; C, sc thigh CSA at baseline and after 12 wk; D, change in sc thigh CSA from baseline to 12 wk, in percent. Bars and error bars represent mean values and SEM, respectively. *, Significant change within a group; (*), P < 0.08).
Isokinetic torque, 1 RM, and leg extensor power
No differences in isokinetic quadriceps torque or leg extensor power were detected between the groups at baseline. Isokinetic quadriceps torque increased significantly and equally, by approximately 18–20% at both 60 and 180°·sec-1, in the two training groups from baseline to 12 wk but remained unchanged in the two nontraining groups (Fig. 4); 1 RM was only performed in the two training groups and increased significantly and equally, by approximately 65%, in both groups. Leg extensor power, measured in the Nottingham Power Rig, increased from baseline to 12 wk in the RT + placebo group (15 ± 3%, P < 0.0078) but did not change in any of the other three groups (Fig. 4).
View larger version (20K):
[in this window]
[in a new window]
Figure 4. Isokinetic quadriceps torques at A (60°·sec-1) and B (180°·sec-1). C, Quadriceps leg power; D, 1 RM leg extension. Measurements performed at baseline and after 12 wk. Bars and error bars represent mean values and SEM, respectively. *, Significant change within a group. See Results for further details.
Fiber types
There were no differences in fiber type composition among the three groups at baseline [type 1, P > 0.34 (Kruskal-Wallis); type 1/2a, P > 0.14; type 2a, P > 0.34; type 2a/2x, P > 0.46; type 2x, P > 0.84] (Fig. 5). No significant changes in fiber type composition were observed in the RT + placebo group from baseline to 12 wk. However, there was a tendency toward a decrease in type 1 fibers from 62.9 ± 6.3% to 52.6 ± 4.7% (P < 0.148), a tendency toward an increase in type 2a fibers from 24.4 ± 4.9% to 34.0 ± 5.6% (P < 0.109), and a tendency toward an increase in type 2a/2x fibers from 2.9 ± 1.2% to 4.9 ± 1.3% (P < 0.148), whereas type 2x fibers were unchanged [8.7 ± 3.1% (baseline) vs. 7.7 ± 3.5%, P > 0.84] (Fig. 5A). The same pattern was observed in the RT + GH group, where type 1 fibers tended to decrease from 61.4 ± 6.7% to 47.2 ± 4.0% (P < 0.109), type 2a fibers increased significantly from 24.2 + 5.5% to 35.4 + 5.5% (P < 0.0313), type 2a/2x increased significantly from 3.2 ± 1.2% to 11.0 ± 2.7% (P < 0.0156), and type 2x fibers were unchanged [9.8 ± 3.2% (baseline) vs. 5.0 ± 1.2%, P < 0.22] (Fig. 5B). In the group receiving only GH, a different pattern was observed, with a tendency for a decrease in type 2a fibers from 34.0 ± 5.5% to 22.4 ± 1.9% (P < 0.078) and a tendency for an increase in type 2x fibers from 7.5 ± 2.7% to 20.0 ± 2.6% (P < 0.078) (Fig. 5C).
View larger version (34K):
[in this window]
[in a new window]
Figure 5. Fiber type and MHC composition in quadriceps muscle biopsies obtained at baseline and after 12 wk. Bars and error bars represent mean and SEM, respectively. *, Significant change within a group. See Results for further details.
Fiber size
There were no differences in type 1 and type 2 fiber size among the three groups at baseline (type 1, P > 0.54, Kruskal-Wallis; type 2, P > 0.59). No significant changes were observed in type 1 and type 2 fiber size in any of the groups from baseline to 12 wk [RT + placebo, type 1, from 4521 ± 451 (baseline) to 4618 ± 364 (12 wk), P > 0.64; RT + placebo, type 2, from 3832 ± 418 to 4262 ± 530, P > 0.19; RT + GH, type 1, from 5066 ± 383 to 5172 ± 301, P > 0.46; RT + GH, type 2, from 4070 ± 234 to 4380 ± 291, P > 0.37; GH alone, type 1, from 4668 ± 492 to 4183 ± 304, P > 0.10; GH alone, type 2, from 4050 ± 241 to 4143 ± 200, P > 0.81] (all areas in µ·m-2). Similarly, no significant changes in fiber size from baseline to 12 wk were detected when the fibers were further divided into type 1, type 1/2a, type 2a, type 2a/2x, and type 2x (data not given). When pooling data from the two training groups, there was still no significant change in type 1 fiber size [from 4775 ± 298 (baseline) to 4877 ± 243 (12 wk), P > 0.42]; whereas the increase in type 2 fiber size almost reached significance (from 3943 ± 242 to 4317 ± 304, P < 0.064).
MHC composition
There were no differences in MHC composition among the three groups at baseline [MHC 1,P > 0.60 (Kruskal-Wallis); MHC 2A, P > 0.47; MHC 2x, P > 0.28] (Fig. 5). MHC composition did not change from baseline to 12 wk in the RT + placebo group (Fig. 5A). In the RT + GH group, MHC 1 decreased from 67.3 ± 5.7% to 53.7 ± 4.0% (P < 0.0234), MHC 2A increased from 26.2 ± 4.7% to 43.4 ± 3.7% (P < 0.0156), and MHC 2x decreased from 6.5 ± 1.9% to 2.9 ± 0.8% (P < 0.0469) (Fig. 5B). In the group receiving only GH, MHC 1 and MHC 2A composition did not change; whereas MHC 2x increased substantially and significantly, from 2.8 ± 1.2% to 11.1 ± 2.3% from baseline to 12 wk (P < 0.0156) (Fig. 5C).
Discussion
In the present study, we found no effect of 12-wk GH administration alone on isokinetic muscle strength or muscle power in healthy elderly men (Fig. 4, A–C). To our knowledge, the isolated effects of GH administration on muscle strength in healthy elderly humans has only been investigated in two previous studies, and they both showed that GH administration alone does not increase muscle strength in either elderly men (26) or women (27). Thus, our findings support the existing evidence. Even though FFM increased significantly with GH administration alone, it is impossible, from DEXA measurements, to obtain information about the various components that constitute the increase in FFM (44). Obviously, this is a major issue in trials investigating effects of GH administration on body composition and, in particular, on muscle mass, because GH is known to cause water retention (9). However, GH administration alone did not increase quadriceps muscle CSA determined by NMRI (Fig. 4) or induce fiber hypertrophy determined from quadriceps muscle biopsies. These findings further support that GH administration alone has little, if any, effect on skeletal muscle hypertrophy and strength in healthy elderly men.
When GH administration was combined with RT, no further improvements in muscle strength were observed in the present study. In addition, both RT + placebo and RT + GH caused a significant, but similar, increase in muscle CSA, determined by NMRI (Fig. 4), whereas no effect was detected on muscle fiber sizes in either group. These findings support previous investigations in healthy elderly (17, 19) and young men (21) as well as in young power athletes (18, 45) but are in contrast to the findings by Welle et al. (20), who found that GH administration actually augmented training-induced strength in healthy elderly men. The reason for the conflicting results obtained by Welle et al. vs. Yarasheski et al. and Taaffe et al. is not immediately intelligible. Obviously, it cannot be excluded that GH administration may have resulted in small increments in muscle strength that escaped detection by the methods we used. Likewise, it cannot be excluded that elite athletes, in whom very small increases in performance result in dramatic improvements in ranking, may benefit from additional GH administration, although this may be difficult to substantiate scientifically. However, from the present study and from the previously published studies, we conclude that elderly men benefit substantially from RT, but additional GH administration does not further improve the gains in muscle hypertrophy and strength.
Several (23, 24, 46), but not all (16, 47), studies investigating GH replacement therapy in GH-deficient adult patients have demonstrated small, but significant, increases in muscle strength. Although elderly are sometimes referred to as GH-deficient, this term may not be appropriate, because GH deficiency involves a heterogeneous group of patients covering both a wide age span and a wide range of hormonal disorders requiring several hormones for replacement therapy (sex steroids, corticosteroids, thyroid hormones, and others). It is therefore easy to imagine potential confounders in studies involving GHD patients, and results obtained in GHD patients are thus difficult to compare with results obtained in healthy elderly subjects.
GH administration alone resulted in a dramatic increase in MHC 2X composition, paralleled by a similar increase in type 2x fibers, which almost reached statistical significance (Fig. 5C). It has been suggested that, with increasing age, there is a trend toward a shift in muscle MHC composition from 2x to 2A and 1 (48). It may thus be hypothesized that GH, by rejuvenating systemic IGF-I levels, may shift MHC composition to a more youthful profile in elderly people. However, in terms of muscle functionality, these changes in muscle MHC composition are not likely to be revealed by the muscle tests performed in the present study (Fig. 4). It would have been more appropriate to measure changes in the rate of force development. To our knowledge, there are no published data on the effect of GH administration on MHC composition in healthy humans. Several animal studies suggest that GH may influence fiber composition, but the results are far from uniform. In one study, GH administration to normal rats was found to selectively induce type II fiber hypertrophy (49). It is known that muscle fibers express IGF-I receptors (50), and current evidence suggests that muscle fibers also express GH receptors (51). However, the potential signaling pathways for the observed effect of isolated GH administration in the present study, either directly or indirectly through the actions of IGF-I, remain to be elucidated.
Skeletal muscle MHC or fiber type composition did not change with RT alone, whereas a decrease in MHC 1 and an increase in MHC 2A, accompanied by similar changes in fiber type composition, were observed when RT was combined with GH (Fig. 5, A and B). The latter findings are consistent with changes reported in previous RT studies in elderly (48) and suggest that GH administration combined with RT does not change the direction of MHC changes that occur with RT alone. However, a recent study demonstrated increased levels of MHC 1 mRNA and decreased levels of MHC 2A and 2x mRNA with RT in elderly subjects (52). Because we demonstrated either no change (RT alone) or even opposite changes in MHC composition at the protein level, it may be that the apparent increase in MHC 1 mRNA and concomitant decreases in MHC 2A and 2x mRNA are not propagated to the translational or posttranslational levels. Alternatively, a change in the protein degradation among the different MHC isoforms may counteract the changes in mRNA expression. Currently, the training stimulus was able to completely overrule the effects of isolated GH administration on MHC changes, and these findings support that mechanical stretch and systemic GH (or IGF-I) act through different signaling pathways. MHC 2x fibers are known to possess low insulin sensitivity (53), and a shift in MHC composition toward higher percentage of MHC 2x may play a role in the well-known effect of GH administration to decrease insulin sensitivity.
GH administration caused substantial changes in body composition, determined by DEXA scanning, with increased FFM and decreased FM after 12 wk (Fig. 1). These findings confirm previous observations (15, 26) and support the widely accepted view that GH plays a major role in determining body composition. The decrease in FM was also evidenced by the decrease in sc thigh CSA, measured by NMRI, in the two groups receiving GH (Fig. 3D). Although GH has a major impact on bone growth, no change in BMC was observed in any of the GH groups. However, the study period may very well be too short to make definitive statements about changes in BMC.
Currently, 12 of the 15 subjects receiving GH and completing the study experienced side effects. A high incidence of side effects has been reported in several papers; but, in retrospect, we obviously targeted a too-high GH dose, and further dose reductions are necessary in future studies involving healthy elderly men.
In conclusion, the present study confirms that GH administration alone does not increase muscle strength or hypertrophy in healthy elderly men. Furthermore, healthy elderly men adapt beneficially to a 12-wk RT program, with substantial increases in muscle mass and muscle size. However, GH administration does not further augment these training-induced adaptations. GH administration alone seems to increase muscle MHC 2x, possibly through rejuvenating systemic IGF-I levels, whereas this response is completely overruled when RT is combined with GH. The present study does not support that GH therapy either alone or in combination with RT should be used to augment muscle strength and/or hypertrophy in elderly men.