This investigation determined the influence of acute and chronic resistance exercise on responses of growth hormone (GH) molecular variants in women. Seventy-four healthy young women (23 ± 3 yr, 167 ± 7 cm, 63.8 ± 9.3 kg, 26.3 ± 4.0% body fat) performed an acute bout of resistance exercise (6 sets of 10 repetition maximum squat). Blood samples were obtained pre- and postexercise. Resulting plasma was fractionated by molecular mass (fraction A, >60 kDa; fraction B, 30–60 kDa; and fraction C, <30 kDa) using chromatography. Fractionated and unfractionated (UF) plasma was then assayed for GH using three different detection systems (monoclonal immunoassay, polyclonal immunoassay, and rat tibial line in vivo bioassay). Subjects were then matched and randomly placed into one of four resistance exercise training groups or a control group for 24 wk. All experimental procedures were repeated on completion of the 24-wk resistance training programs. After acute exercise, immunoassays showed consistent increases in UF GH samples and fractions B and C; increases in fraction A using immunoassay were seen only in the monoclonal assay. No consistent changes in bioactive GH were found following acute exercise. Conversely, chronic exercise induced no consistent changes in immunoassayable GH of various molecular masses, whereas, in general, bioassayable GH increased. In summary, although acute exercise increased only immunoactive GH, chronic physical training increased the biological activity of circulating GH molecular variants. Increased bioactive GH was observed across all fractions and training regimens, suggesting that chronic resistance exercise increased a spectrum of GH molecules that may be necessary for the multitude of somatogenic and metabolic actions of GH.
- strength training
growth hormone (GH) is considered a “family of hormones,” as more than 100 different variant forms are known to exist in the circulation (2). These variants include 22-kDa GH monomers (the most frequently studied form), 20-kDa mRNA splice variants, disulfide-linked homodimers and heterodimers, glycosylated GH, high molecular mass oligomers, GH bound to GH-binding protein, and hormone fragments (e.g., 5 and 17 kDa) resulting from proteolysis. The specific biological activity of each of these GH variants has yet to be fully elucidated; however, it is clear that concentrations of these GH variants are dramatically affected by exercise (11, 32). Study of the complexities of the molecular forms of GH in human blood in response to exercise, however, is confounded by the use of numerous assay detection systems (e.g., radioimmunoassays, enzyme-linked immunoabsorbent assays, and in vitro and in vivo bioassays) and by various combinations of acute exercise program variables (e.g., intensity, rest period lengths, total work) that modify the GH response to exercise (13, 14, 16).
In early work by our research group, we demonstrated that different combinations of acute exercise program variables create dramatically different responses in postexercise concentrations of immunoreactive GH in men and women (13, 14, 16). Previously (16), we carefully controlled and isolated the effects of load [5 vs. 10 repetition maximum (RM)], interset rest period length (1 vs. 3 min), and total work. Maximal resistance exercise-induced immunoreactive GH responses were observed during the protocol that involved moderate loads (10 RM), short rest periods (1 min), and high total work. Importantly, not all combinations of these variables produced significant increases in GH. Therefore, the acute exercise program variables are crucial mediators of the GH response to resistance exercise.
More recently, using women, we documented that acute resistance exercise-induced GH responses were dependent on the GH molecular mass fraction examined and the assay detection system used (11). In that study, we concluded that acute resistance exercise may specifically increase release of disulfide-linked GH dimers. Furthermore, most of the GH released after exercise was able to dimerize the GH receptor in vitro, suggesting that these molecules had the two intact binding sites required to initiate signal transduction in target cells. These findings were important, as they suggested that acute exercise could increase the biological activity of GH by inducing the release of molecular forms with extended half-lives, thereby sustaining biological action.
Few data exist on the effects of chronic resistance training on resting or exercise-induced GH concentrations. Previous investigations have shown no difference in resting GH concentrations in responses to chronic (8–24 wk) resistance exercise in healthy young men and women (3, 19, 20, 22, 30). Studies comparing resistance exercise-induced GH concentrations before and after long-term resistance exercise have shown no change (1, 19, 22) or a slight increase (3). Importantly, however, none of these studies has considered multiple assay detection systems or investigated differential responses of molecular mass variants while studying GH responses to long-term resistance exercise. Therefore, the purpose of this investigation was to examine changes in resting and resistance exercise-induced responses of different GH molecular variants measured via multiple assay detection systems to different long-term resistance training programs. We hypothesized that GH responses to chronic resistance exercise are dependent on 1) the specific resistance training program used (i.e., “hypertrophy” or “strength” programs), 2) the molecular mass variant studied, and 3) the assay detection system employed.
A between-subjects design was used to determine the effects of 24 wk of resistance training on molecular mass variants of GH. Participants performed an acute bout of resistance exercise with associated preexercise and postexercise blood draws. All experimental procedures were subsequently repeated after 24 wk of chronic resistance exercise training. Plasma samples obtained during the study were fractionated using a sizing column and subsequently assayed for GH using two different immunoassays and the rat tibial line in vivo bioassay.
Seventy-four previously untrained women (means ± SD physical characteristics: 23 ± 3 yr, 167 ± 7 cm, 63.8 ± 9.3 kg, 26.3 ± 4.0% body fat) volunteered to participate in this investigation, which was approved by the Pennsylvania State University Human Use and Review Board, the Institutional Animal Care and Use Committee, and the Human Use Review Office of the United States Army Surgeon General. The risks and benefits of the investigation were thoroughly explained to all human participants, and written informed consent was subsequently obtained.
Exclusionary criteria for this study included smoking, recreational drug use, and pregnancy. Once enrolled, all participants were screened for nutritional habits by a registered dietician 2 wk before initial testing to assure normal dietary intakes and eliminate confounding influences of aberrant diets. Furthermore, participants were found to have habitually consumed adequate energy and macronutrient composition (30–40% energy from fat, 15–20% energy from protein, 40–55% energy from carbohydrate). For menstrual status, we used the same procedures as in our prior work on this topic of GH responses (11). Thus participants were menstruating regularly and defined as eumenorrheic (28- to 32-day menstrual cycles during the previous year); all testing was performed during the follicular phase of the menstrual cycle. None of the subjects was on any birth control. Although enrolled subjects were considered healthy and active, none had performed a structured physical conditioning program for the previous 12 mo.
Participants were initially matched for age, height, body mass, one-repetition maximum (1 RM) squat and bench press performance (see below), and physical activity history. Then, participants were randomly placed in either a resistance training group or a nonexercise control group in a 2:1 ratio. Before initiation of data collection, all individuals participated in a 2-wk familiarization period to accustom them to testing and training procedures.
Back squat and bench press 1 RM strength was measured using the Plyometric Power System (PPS; Norsearch, Lismore, Australia), previously described in detail (33). The PPS utilized a modified Life Fitness (Franklin Park, IL) Smith Machine that allowed only vertical translation of the bar. Linear bearings attached to either side of the bar allowed vertical movement along two steel shafts with minimal friction. For the squat, each participant descended to the parallel position (by flexing the knees and hips until the greater trochanter of the femur reached the same horizontal plane as the superior border of the patella) and, on a verbal signal from the tester, ascended to the upright starting position while maintaining proper form and technique throughout the lift. For the bench press, each participant lowered the bar until contact with the chest was achieved and subsequently lifted the bar back to full elbow extension. Any trials that failed to meet the standardized technique criteria were discarded.
Acute heavy resistance exercise testing protocol.
The acute heavy resistance exercise testing (AHRET) protocol consisted of six sets of 10 RM squats with 2-min interset rest periods. Initially, the 10 RM load was calculated as ∼75% of the subject's 1 RM squat. If the subject failed to perform the 10 repetitions because of fatigue on any given set, the load was immediately adjusted to permit completion of the remaining repetitions. The time required to complete the entire AHRET was ∼15–20 min. This test was performed both before and after 24 wk of training. All exercise tests were performed between 0630 and 1100 (24-h clock) after an 8- to 12-h fast, and the time of day for each subject during the pre- and posttesting was within an hour.
Resistance training protocol.
We have previously described the conditioning programs used in the present study (17). In brief, all individuals (with the exception of the control group) participated in a 24-wk periodized resistance training program performed on 3 alternating days/wk. The four resistance training protocols were as follows: 1) a total body high-intensity/low-volume (3–8 repetitions/set) resistance exercise routine designed to maximize increases in strength (i.e., TB Strength; n = 12), 2) a total body moderate-intensity/moderate-volume (8–12 repetitions/set) resistance exercise routine designed to maximize increases in muscular hypertrophy (i.e., TB Hypertrophy; n = 15), 3) an upper-body-only high-intensity/low-volume (3–8 repetitions/set) resistance exercise routine designed to maximize increases in upper-body strength (i.e., UB Strength; n = 18), and 4) an upper-body-only moderate-intensity/moderate-volume (8–12 repetitions/set) resistance exercise routine designed to maximize increases in upper-body muscular hypertrophy (i.e., UB Hypertrophy; n = 18). We included both total body and upper-body-only resistance training, because previous research (10) has shown that acute hormonal responses and subsequent strength adaptations to chronic training are greater during total body than upper-body-only resistance exercise.
The 24-wk program consisted of free weight and machine exercises and was divided into two 12-wk mesocycles, each consisting of three short mircocycles (see Table 1 for details). Furthermore, all training groups participated in a standard aerobic conditioning program (25–35 min at 70–85% of maximum heart rate) 3 days/wk. This aerobic conditioning protocol has been shown to have a negligible impact on strength and power development (31). All sessions were individually supervised by certified strength and conditioning specialists who directly monitored all training sessions to optimize the training adaptations (21).
We have previously published the changes in performance (17) and muscle size (18) that occurred in subjects using these programs, but, because of the menstrual phase criteria, some women were excluded from this study's findings. The findings on our groups indicated that arm cross-sectional area (CSA) increased at week 12 (∼11%) and week 24 (∼6%) in all training groups, and thigh CSA increased at week 12 (∼3%) and week 24 (∼4.5%) only in TB Strength and TB Hypertrophy groups. Squat 1 RM increased at week 12 (∼24%) and week 24 (∼11.5%) only in TB Strength and TB Hypertrophy, and all training groups increased 1 RM bench press at week 12 (∼16.5) and week 24 (∼12.4%). Increase in the AHRET total work was ∼12% for TB Strength and TB Hypertrophy and ∼8% for the other two upper-body groups, with improvements in strength and tolerance to exercise stress.
Before each AHRET session, each subject rested quietly in the laboratory for 30 min before we obtained the preexercise blood sample to reduce any anticipatory responses in hormonal concentrations. Preexercise blood samples were obtained 15 min before AHRET via standard venipuncture. A second sample of whole blood was obtained immediately after (i.e., within 2 min) completion of AHRET.
For this investigation, time points for individual blood draws were operationally defined as follows: T1, week 0, preexercise blood draw; T2, week 0, postexercise blood draw; T3, week 24, preexercise blood draw; T4, week 24, postexercise blood draw.
Therefore, for this investigation, we were able to 1) determine the effects of acute resistance exercise in untrained women (i.e., T1 vs. T2), 2) determine the effects of acute resistance exercise following 24 wk of periodized resistance training (i.e., T3 vs. T4), 3) determine the effects of chronic resistance exercise training on resting GH concentrations (T1 vs. T3), and 4) determine the effects of chronic resistance exercise on resistance exercise-induced GH concentrations (T2 vs. T4).
Blood processing and fractionation.
Plasma was obtained from each sample by centrifugation (800 g, 10 min) and distributed according to the scheme depicted in Fig. 1.
Fifteen milliliters of plasma were fractionated on a 100-cm-long Sephacryl S-100 HR sizing column (26-mm id), and the resulting 100 tubes were pooled into three larger fractions containing molecules with apparent molecular masses >60 kDa (i.e., “fraction A”), 30–60 kDa (i.e., “fraction B”), and <30 kDa (i.e., “fraction C”). Two Sephacryl columns were used throughout the course of this study: one for processing each of the 60 pre-AHRET plasma samples and the other for the 60 post-AHRET samples. Each of the two columns was calibrated using blue dextran (mol wt 2,000,000), BSA (mol wt 66,000), carbonic anhydrase (mol wt 29,000), cytochrome c (mol wt 12,400), and aprotinin (mol wt 6,500), provided in a molecular weight standards kit (Pharmacia, Uppsala, Sweden). Columns were washed with 0.05 M NH4HCO3, pH 8.0 (1 column volume), between processing of each plasma sample and were recalibrated after 15 plasma samples had been processed. Regression analysis of these calibration curves [log(mol wt) vs. the ratio of eluted protein to column void volume (Ve/Vo)] yielded lines of the form y = −1.069x + 6.1144 and y = −1.0196x + 6.0595. Correlation coefficients for these regressions were 0.997 and 0.996, respectively.
Circulating plasma human GH was determined with the use of three different immunoassays: 1) National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) competitive radioimmunoassay (NIDDK RIA), 2) the Nichols Institute Diagnostics immunoradiometric assay (Nichols IRMA; San Juan Capistrano, CA), and 3) the rat tibial line in vivo bioassay. All assays were validated with respect to linearity, parallelism, and recovery. Log-logit and log-log standard curve fitting regression were used for the NIDDK RIA and Nichols IRMA, respectively. To eliminate interassay variance, all samples from a subject were assayed within the same batch using a gamma counter and curve-fitting algorithms (EG&G Wallac Gamma Counter, Turku, Finland) for the RIAs. All assays produced variances <10%; significant changes in GH concentrations reported in results exceed this variance.
In each case, recoveries of immunoreactive GH after chromatography were 1.5- to 3.0-fold higher than GH concentrations measured in unfractionated plasma. High recoveries of GH after column chromatography have been reported by other investigators (4); the reason(s) is unknown but may reflect removal of an inhibitory substance(s) in the starting material.
NIDDK polyclonal RIA.
The NIDDK RIA reagents were obtained from Dr. A. F. Parlow and the National Hormone and Pituitary Program of the NIDDK. The GH antigen (NIDDK-GH-I-3; AFP-11019B) was iodinated on-site. The primary polyclonal antibody was rabbit NIDDK-anti-GH-2. The biological potency of the GH reference preparation (NIDDK-GH-RP-1; AFP-4793B) was 2.2 international units (IU)/mg as determined in the hypophysectomized female rat body weight gain bioassay. This reference preparation was serially diluted from 25 to 0.195 ng/ml (8 different standard concentrations); these standard concentrations were subsequently used to create the standard curve in both immunoassays. The secondary antibody was sheep anti-rabbit globulin, used in conjunction with polyethylene glycol (PEG). After centrifugation, tubes were aspirated, and the pellet was counted in a gamma counter for 120 s (EG&G Wallac Gamma Counter). The sensitivity for this assay, using the B0 ± 2 SD method, was 0.10 ng/ml. Intra-assay variances for low, medium, and high GH concentrations were <10.0%.
This commercially available assay used two monoclonal antibodies of high affinity and specificity for GH; each detects a different epitope on the GH molecule. One of the antibodies was labeled for detection, while the other was coupled to biotin. The sensitivity for this assay, using the B0 ± 2 SD method, was 0.04 ng/ml. Intra-assay variances for low, medium, and high GH concentrations were 7.2, 5.2, and 5.4%, respectively.
Rat tibial line in vivo bioassay.
Concentrations of biologically active GH in unfractionated and fractionated plasma samples were determined according to the method of Greenspan et al. (8). Briefly, female Sprague-Dawley hypophysectomized rats (Hilltop Labs, Scottsdale, PA; hypophysectomized at 26–28 days of age) were used 2 wk after surgery. Animals that weighed <80 or >100 g at the time of sample injection were excluded. The following criteria were taken as evidence for completeness of hypophosectomy: failure to gain >7 g in the 10 days following the operation, deterioration of body tonus, maintenance of infantile (“smooth”) hair, and absence of pituitary remnants in the sella turcica at autopsy. Animals were injected subcutaneously once daily for 4 days with either 1) experimental plasma samples, 2) a standard GH preparation (United States Department of Agriculture bovine GH B-1 AFP 5200, 1.4 IU/mg at total doses of 10, 30, or 90 μg), or 3) physiological saline (control). Twenty-four hours after the last injection, the animals were killed, tibial epiphyseal plates were stained, and plate widths were measured in double-blind fashion using an ocular micrometer (10 readings averaged across the plate width for each sample). GH responses were expressed in terms of a purified human pituitary preparation (3.0 IU/mg). Assay variance was 7.2%. Average tibial widths of animals injected with bovine GH standard were as follows: saline, 147 ± 11 μm (means ± SD); 10 μg, 168 ± 14 μm; 30 μg, 190 ± 24 μm; 90 μg, 233 ± 22 μm (y = 1.0768x − 166.04, r2 = 0.96).
All results are reported as means ± SE. A 5 × 2 × 2 (group × time × test) ANOVA with repeated measures on the test and time variables was used to determine differences in GH concentrations among the training groups after acute (pre- and post-AHRET) and chronic (pre- and post-24 wk of chronic exercise training) exercise. An ANOVA was conducted for each assay condition: NIDDK RIA, Nichols IRMA, and the tibia line in vivo bioassay. In the case of a significant F-score, a Fishers least significant difference (LSD) post hoc test was performed to determine where significant differences lie. An α-level of P ≤ 0.05 was used for statistical significance.
Exercise-Induced Changes in GH
Plasma concentrations of GH fractions for the individual groups as measured via the NIDDK RIA are shown in Fig. 2. This assay showed very consistent responses among the groups. For each group, acute exercise evoked a significant (P < 0.05) increase in GH in unfractionated plasma and in fractions B and C; this response was apparent both before and after the 24-wk training protocol. After chronic training, a greater increase in postexercise unfractionated GH, compared with the pretraining, postexercise, unfractionated GH, was noted in all training groups except for TB Hypertrophy. Alternately, the control group showed no changes in postexercise GH responses at 24 wk of training. The NIDDK assay was unable to detect significant changes in fraction A at any time point.
Results for the Nichols IRMA are shown in Fig. 3. The Nichols IRMA showed that acute exercise increased concentrations of GH in unfractionated plasma and all three GH molecular mass variants. This pattern was similar before and after chronic resistance training. Preexercise GH concentrations did not change after 24 wk of training (T1 vs. T3); however, postexercise GH concentrations following training (T2 vs. T4) showed inconsistent patterns between fractions and training groups. The control group showed no changes in GH concentrations after 24 wk.
Biologically active GH concentrations (as measured by the rat tibial line in vivo bioassay) are shown in Fig. 4. Acute resistance exercise failed to induce consistent changes in various fractions of biologically active GH. However, chronic resistance exercise training increased preexercise (T1 vs. T3) and postexercise (T2 vs. T4) GH for most, but not all, molecular fractions and training groups. Similar to the immunoassay results, the control group did not increase biologically active GH after 24 wk.
Summaries of the effects of exercise on circulating concentrations of molecular mass variants of GH are presented in Table 2 (acute exercise) and Table 3 (chronic exercise). Together, Tables 2 and 3 indicate that increases in GH after acute exercise were more readily observable with the use of conventional assays (NIDDK RIA and Nichols IRMA). Alternately, compared with the corresponding pretraining time points, increases in GH after 24 wk of resistance training were more readily observable using the rat tibial line in vivo bioassay.
Comparisons of Molecular Mass Variants
When molecular mass variants were assayed using the NIDDK RIA, there was a main effect for molecular mass variants (unfractionated > fraction C > fraction B > fraction A). With the use of the Nichols IRMA, there was no difference between the concentrations of fractions C and B; however, the concentrations of both these variants were greater than for fraction A (unfractionated > fraction C = fraction B > fraction A). Alternately, using the rat tibial in vivo bioassay, unfractionated plasma had the highest concentration; however, there was no main effect for molecular mass variants (unfractionated > fraction C = fraction B = fraction A; see Figs. 2–4).
The salient discoveries in this investigation were, first, that changes in GH concentrations following acute and chronic resistance exercise were dependent on the assay used. In general, acute exercise stimulated increases in forms of GH that were capable of generating an immunoreactive response but not an increase in biological activity. Conversely, chronic resistance exercise (24 wk) induced no change in immunoreactivity but increased circulating concentrations of biologically active GH. Increased GH biological activity following chronic physical activity is an important and novel finding and provides a possible mechanism for favorable somatogenic adaptations after chronic exercise training.
The second major finding was that small (<30 kDa) molecular mass variants generated the largest immunoreactive response; however, large (>60 kDa) molecular mass variants were equally as potent as small (<30 kDa) and medium (30–60 kDa) fractions for generating a biological response. This indicates the importance of GH aggregates and/or GH bound to GH-binding protein for somatogenic adaptations.
Acute Resistance Exercise
Following acute resistance exercise (6 sets of 10 RM squats), there was a significant increase in immunoreactive, but not bioactive, GH in women (summarized in Table 2). These results closely agree with our previous findings in women (11). Using the polyclonal NIDDK assay, we found an acute increase in all GH variants except for fraction A (>60 kDa). Alternately, the monoclonal Nichols assay detected increases in all variants, suggesting that GH oligomers and/or GH bound to GH-binding protein may be better detected by the monoclonal immunoassay. Comparison of the two immunoassays revealed that the monoclonal antibody used in the Nichols IRMA resulted in greater unfractionated GH concentrations than the polyclonal antibody used in the NIDDK RIA. A similar finding was reported in our previous study (9). The difference between monoclonal and polyclonal antibodies in GH detection may be a result of interfering molecules; however, it is currently unknown why the Nichols IRMA produces higher concentrations than the polyclonal NIDDK RIA. Because of the relative design of this study, it is possible that total work interacted with the intensity of the AHRET to affect the postexercise response, but it would not impact the changes observed in resting concentrations. The ability to perform more work is a vital marker of improved function that results in a potentially greater endocrine response, as in this case. However, no systematically apparent patterns of changes were observed between the exercise training groups as related to volume of work (see Tables 2 and 3).
Few investigations have considered the molecular heterogeneity of GH within exercise paradigms. Wallace et al. (32) employed seven different assays to measure GH and reported postexercise increases following aerobic exercise in all seven. Nindl et al. (24) compared immunoreactive vs. immunofunctional assay results for pulsatile GH after resistance exercise in men and reported quantitative, but not qualitative, differences. Comparison of results across studies is difficult because of differences in techniques used to assay GH isoforms, in sex of subjects, and in type/intensity of exercise testing. In general, however, the results from Wallace et al. (32), Nindl et al. (26), and Hymer et al. (11) agree closely with those of the present study: exercise is a potent stimulus for immunoassayable GH of most (if not all) variants.
Increases in immunoreactive GH activity were dramatically contrasted with a lack of change in bioactive GH in response to acute exercise. This finding agrees with previous work by our group (11) that showed no significant increases in bioactive GH in women in response to acute exercise using the same exercise protocol. Differences between immunoreactive and bioactive GH release may be explained by the heterogeneity of GH-producing cells in the anterior pituitary. Prior research has shown that low-density (LD) somatotrophs have fewer secretory granules, and these cells preferentially secrete monomeric 22-kDa GH (6, 29). Conversely, high-density (HD) somatotrophs are densely packed with granules, and these cells secrete a high percentage of polymeric GH aggregates (6, 29). These aggregates may be more somatogenic than monomers (5); however, GH aggregates have low immunoreactivity. Therefore, perhaps the HD somatotrophs in women are less responsive to the acute stress of resistance exercise.
The current results and those of Hymer et al. (11) are in stark disagreement with those of McCall et al. (23), who found an increase in bioassayable, but not immunoassayable, GH in men following acute physical activity. These dissimilarities could be attributed to the vastly different exercise protocols used (6 sets of 10 RM squats vs. unilateral plantar flexion) or to sex differences in GH physiology. The latter is supported by Jaffe et al. (12), who have previously shown that GH regulation and secretory patterns are sexually dimorphic. Thus it is feasible that other sex differences in GH physiology may exist. However, the greater physiological stress associated with the squat protocol would significantly increase cortisol concentrations compared with the plantar flexion protocol, which may inhibit the somatogenic influence of GH in the rat tibial line in vivo bioassay. Additionally, previous work by our group (25) compared sexes in GH responses to acute resistance exercise. In that study, no differences between sexes were found in GH concentrations of immunoassayable GH and immunofunctional GH (which is GH that has both binding sites available for receptor interaction). Therefore, it seems likely that the differences between the current results and those of McCall et al. (23) are due to differences in the exercise protocol used.
Chronic Resistance Exercise
Following the 24-wk supervised resistance exercise training protocols, there were very few changes in resting immunoreactive GH variants over pretraining time points; however, a general increase in resting bioactive GH was noted across specific variants and training protocols (summarized in Table 3). These results agree with previous studies that showed no long-term training-induced changes in immunoassayable GH (3, 19, 20, 22, 23). Notably, however, our results suggest that training increased bioactive GH across fractions and exercise regimens; this is a physiologically important and novel finding.
Although 22-kDa GH is the most studied isoform because of its somatogenic (e.g., increased bone growth and lean mass) and metabolic (e.g., enhanced lipolysis) effects, other molecular mass variants have important physiological roles as well. Increased circulating GH aggregates and fragments may be the underlying mechanism whereby chronic exercise increases GH bioactivity. As stated previously, GH aggregates may be more somatogenic than monomers (5). Furthermore, research has shown that GH fragments have potent metabolic effects. The fragment possessing amino acids 1–43, which is found in humans, has insulin-potentiating properties in mice (7, 28); alternately, the amino acid 44–191 fragment is quite diabetogenic (27). Thus stimulation of both aggregates (e.g., dimers, trimers, and oligomers) and fragments (e.g., 5 and 17 kDa) following long-term resistance exercise may lead to more effective interaction with the wide array of target tissues susceptible to the actions of GH, such as muscle, bone, and immune cells. It might be hypothesized that a more diverse spectrum of GH molecular variants differentially interact with a host of tissues that are affected by GH in either somatogenic or metabolic functions.
Molecular Mass Variants
Small (<30 kDa) molecular mass variants generated the largest immunoreactive response. This finding is not surprising, because 22-kDa GH is typically used as the antigen to raise monoclonal and polyclonal antibodies. Thus the antibodies used during immunoassays are more specific to small molecular mass fractions. However, large (>60 kDa) molecular mass variants were equally as potent as small (<30 kDa) and medium (30–60 kDa) fractions in generating a biological response. Similarly, Ellis et al. (5) concluded that GH aggregates are more biologically active than monomers. Our findings indicate 1) that 22-kDa GH, the most commonly studied form, is not the sole factor in determining GH action and 2) the importance of GH aggregates and/or GH bound to GH-binding protein for somatogenic adaptations.
The findings of the current investigation have distinct implications for the future study of GH molecular mass variants. For instance, Kraemer et al. (15) and Hakkinen et al. (9) have shown that GH responses to acute resistance exercise are reduced in old compared with young humans. This finding may provide, in part, a mechanism for the sarcopenia that occurs with aging. Interestingly, however, differentiating GH responses according to molecular mass variant and biological activity have yet to be studied in this population. Similarly, direct gender comparisons of GH physiology using similar analytic techniques have not, to date, been studied. Obviously, many avenues of future research are necessary to complete our understanding of the intricate nature of GH physiology.
Resistance Exercise Groups
As expected, the GH response patterns to exercise were similar among training groups at week 0. In general, posttraining responses were similar as well. The lack of consistent differences among training groups is likely due to the fact that the acute exercise test (6 sets of 10 RM squats) was the same for all training groups and because all training groups performed ∼3 h of resistance training/week. Previous studies have shown that acute exercise-induced increases in GH are affected by the total amount of muscle tissue stimulated, rest intervals, and exercise volume and intensity (13, 14, 16); however, perhaps chronic changes in GH concentrations may be more dependent on the total amount of resistance training. Future research on the importance of acute program variables in increasing bioactive GH following long-term resistance training is necessary.
Because of the multitude of independent variables (assay detection system, GH molecular fraction, acute exercise, chronic exercise, training protocol) used in the study design, it is difficult to discuss more than general patterns in GH responses. For instance, we found increases in certain GH isoforms following specific exercise regimens, whereas other GH isoforms may have decreased. This study clearly demonstrates the complexity of GH physiology. Future studies must consider these complexities to adequately describe the exact nature of GH responses to exercise and the specific biological importance.
In summary, we investigated the changes in GH concentrations of various molecular masses in response to acute and chronic exercise in women. The most important finding was the differential responses of immunoassayable vs. bioassayable GH concentrations following acute and chronic resistance exercise. Acute exercise increased only immunoassayable GH. Alternatively, chronic resistance exercise did not affect immunoassayable GH concentrations despite significant increases in the biological activity of GH isoforms. These findings are novel, as no study has examined GH molecular mass variants following long-term physical training using multiple detection systems. Many details of exercise-induced GH responses are lacking in our understanding and interpretation of GH changes with exercise and physical training. This can be attributed in part to the complexities involved in measuring circulating GH, which include 1) the epitope specificities of the different antibodies used in various immunoassays, 2) the wide spectrum of different molecular GH isoforms circulating in human plasma, 3) the fact that some GH circulates bound to GH-binding protein, and 4) the well-described pulsatile nature of GH release. The heterogeneity of GH action on different target tissues further contributes to that complexity. Nevertheless, the current study provides a novel and multifaceted perspective of GH interactions with physical training.
The views, opinions, and/or findings contained in this report are those of the authors and should not be construed as official Department of the Army position, policy, or decision. Human test participants participated in this study after giving their free and informed voluntary consent. Investigators adhered to AR 70-25 and USAMRDC Regulation 70-25 regarding the use of volunteers in research.
This study was supported by a grant from the Department of Defense, Women's Health Initiative (United States Army Grant DAMD 17-95-5069 to W. J. Kraemer).
We gratefully acknowledge the following: Jana Peters for help in the development of the NIDDK RIA, the research and medical personnel from all the institutions involved with this study, a very large group of research assistants and personal trainers, and the athletic clubs in State College, PA. We also are indebted to the dedicated group of women who participated in this study and made it all possible.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 by American Physiological Society