Christer Malm has part-time employment at the non-profit organization (an economical union) Winternet, Boden, Sweden since 2001. Some of the work for the manuscript was executed using equipment at Winternet. All data, ethical permission, potential patents etc. are filed under Sports Medicine Unit, Umeå University (EPN nr 08–145M). Thus, all past and present data belong to Umeå University. This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials.
Conceived and designed the experiments: CM AE YT PB. Performed the experiments: CM PS AE YT PB. Analyzed the data: JGY CM PS AE YT PB. Contributed reagents/materials/analysis tools: CM PS. Contributed to the writing of the manuscript: JGY CM PS AE YT PB.
The effects of long-term (over several years) anabolic androgen steroids (AAS) administration on human skeletal muscle are still unclear. In this study, seventeen strength training athletes were recruited and individually interviewed regarding self-administration of banned substances. Ten subjects admitted having taken AAS or AAS derivatives for the past 5 to 15 years (Doped) and the dosage and type of banned substances were recorded. The remaining seven subjects testified to having never used any banned substances (Clean). For all subjects, maximal muscle strength and body composition were tested, and biopsies from the vastus lateralis muscle were obtained. Using histochemistry and immunohistochemistry (IHC), muscle biopsies were evaluated for morphology including fiber type composition, fiber size, capillary variables and myonuclei. Compared with the Clean athletes, the Doped athletes had significantly higher lean leg mass, capillary per fibre and myonuclei per fiber. In contrast, the Doped athletes had significantly lower absolute value in maximal squat force and relative values in maximal squat force (relative to lean body mass, to lean leg mass and to muscle fiber area). Using multivariate statistics, an orthogonal projection of latent structure discriminant analysis (OPLS-DA) model was established, in which the maximal squat force relative to muscle mass and the maximal squat force relative to fiber area, together with capillary density and nuclei density were the most important variables for separating Doped from the Clean athletes (regression = 0.93 and prediction = 0.92, p<0.0001). In Doped athletes, AAS dose-dependent increases were observed in lean body mass, muscle fiber area, capillary density and myonuclei density. In conclusion, long term AAS supplementation led to increases in lean leg mass, muscle fiber size and a parallel improvement in muscle strength, and all were dose-dependent. Administration of AAS may induce sustained morphological changes in human skeletal muscle, leading to physical performance enhancement.
Testosterone and other anabolic androgen steroids (AAS) are used by increasing population of professional and recreational athletes with the intention to increase muscle size and improve muscle strength
Short term AAS administration has been shown to induce muscle strength enhancement. The increased muscle strength has been attributed to increased muscle mass which was associated with muscle fiber hypertrophy of both type I and type II fibers
In anti-doping campaign, blood and urine samples are the major materials to be tested
It has been proposed that the effects of AAS on muscle are dose-dependent
The present study will investigate the effects of long term supplementation of AAS on muscle strength and morphology, and explore the relationships between AAS dosage, muscle strength and morphology in elite athletes. We proposed that strength training athletes using AAS will have a higher enhancement in muscle strength through morphological adaptations compared with strength training athletes without using AAS. In addition, the effects of long term AAS supplementation on skeletal muscles will be dose-dependent. Thus, the muscular responses to long term AAS supplementation can be detected and used to separate Doped from Clean athletes.
All participants were informed about the design of the study and written informed consent was obtained from all participants. The study protocol was approved by the Ethics Committee for northern Sweden at Umeå University. This was not an intervention study and no actions were taken to influence the participants' exercise training regime, diet, AAS administration or other activities. Manuscript data is confidential and protected by the Swedish personal integrity law (Personuppgiftslagen 1998:204) and the permission from the ethical board for northern Sweden (
To investigate the long term effects of AAS supplementation on athletes, we recruited 17 strength training elite athletes through personal contact. All subjects were individually interviewed regarding doping substances, physical activity, smoking habits, known illnesses and medication intake. Ten were current users of AAS or AAS derivatives (Doped; age 41.1±8.0 years) and seven reported that they had never used AAS (Clean; age 29.1±6.2 years). Clean subjects had signed a contract with their local clubs and the Swedish Power Lifting Federation, committing them to never use any drugs, under sever monetary punishment. The subjects have been continuously doping-tested with negative results.
The ten Doped subjects were asked to report all banned-substances including doses and intervals taken for the past years. Detailed information of the banned substances and dosage is shown in
Subject | Substances and dosage in recent 5 years | Substances and dosage >5 years ago |
D1 | Testosterone (1250 mg w−1) Dianabol (8 mg d−1) Insulin (10–12 IU d−1) IGF I (50 µg d−1) | Testosterone (1250 mg w−1)Dianabol (8 mg day−1) Trenbolone (262.5 mg w−1) |
D2 | Testosterone (2000 mg w−1) Deca-durabolin (600–800 mg w−1) Dianabol (50 mg d−1) Insulin (12 IU d−1) Ephedrine (60 mg d−1) | Testosterone (2500 mg w−1) Insulin (18 IU day−1) |
D3 | Testosterone (1500 mg w−1) Deca-durabolin (800–1600 mg w−1) Boldone (500 mg w−1) Ephedrine (4–6 IU d−1, 6 days w−1) | Testosterone (1500 mg w−1) Deca-durabolin (600 mg w−1) |
D4 | Testosterone 500 mg w−1 GH (Somatropin) (4–6 IU d−1, 6 days w−1) Deca-durabolin (600 mg w−1) | Testosterone (500 mg w−1) Deca-durabolin (600 mg w−1) |
D5 | Testosterone (500 mg w−1) Deca-durabolin (250 mg w−1) Dianabol (175–350 mg w−1) Ephedrine (10000 IU total) | Testosterone (500 mg w−1) Deca-durabolin (250 mg w−1) Trenbolone (75 mg w−1) |
D6 | Testosterone (500 mg w−1) Deca-durabolin (200 mg w−1) Dianabol (200 mg w−1) | Testosterone (500 mg w−1) Deca-durabolin (200 mg w−1) |
D7 | Testosterone (250 mg w−1) Dianabol (175 mg w−1) | Testosterone (250 mg w−1) |
D8 | Testosterone (250 mg w−1) Deca-durabolin (200 mg w−1) Dianabol (200 mg w−1) Oxar (175 mg w−1) | Testosterone (250 mg w−1) Deca-durabolin (200 mg w−1) Oxar (175 mg w−1) |
D9 | Testosterone (1000 mg w−1) Boldone (1000 mg w−1) Dianabol (105 mg w−1) | Testosterone (1000 mg w−1) Boldone (250 mg w−1) |
D10 | Testosterone (500 mg w−1) Deca-durabolin (400 mg w−1) Trenbolone (150 mg w−1) Dianabol (150–200 mg w−1) | Testosterone (500 mg w−1) Deca-durabolin (400 mg w−1) |
All the subjects reported that they had trained regularly between 4–6 times per week for at least five years. The physical training was defined as self-reported mean hours of exercise training each week during the past five years, and consisted mainly of high intensity resistance training. The Doped group consists of a mixture of bodybuilders, strongmen competitors and weightlifters whereas the Clean group consists of weightlifters only. The mode of resistance training differs slightly between the two groups; the Doped group used both 1–4 repetitions/set and 8–12 repetitions/set, while the Clean group used mainly 1–4 repetitions/set. We have to emphasize that this is the only ethically feasible approach to study long term effects of AAS abuse on athletes.
Subjects performed static squats at a 105° knee angle in a custom-made Smith squat machine and ground reaction forces were recorded by AMTI force plates (464 × 508 mm, Advanced Mechanical Technology Incorporated, Massachusetts, USA). Forces were recorded in x, y and z directions at 100 Hz using the Qualisys Track Manager (QTM) software (Qualisys AB, Gothenburg, Sweden). Maximal force (one single recording), Mean of the highest 50 recordings and Mean of 0.1 sec highest recordings according to rank (highest to lowest) were used for statistical analysis. Personal records (PR) from competition (without tight suits) or equivalent (not all participants had competed in all disciplines) for Bench press, Squat lift and Deadlift were also used for comparisons.
Blood sample of 10 ml was collected from all subjects the same time in the morning after overnight fasting by venipuncture from the cubital vein. Because we could not perform regular doping tests on the subjects and the Doped subjects were not on a “cycle”, indirect indicator of blood hormone level was used to prove/disprove the use of AAS.
Skeletal muscle biopsies were obtained from the vastus lateralis muscle using standard needle or forceps biopsy technique
Serial muscle cross-sections were cut at −20°C by using a Reichert Jung cryostat (Leica, Nussloch, Germany). Eight μm thick sections were stained with haematoxylin-eosin and a modified Gomori trichrome staining for basic histopathology including detection of degenerative processes and inflammation
Five μm thick transverse sections were processed for IHC with different and previously characterized antibodies. For fiber phenotype type classification, serial sections were stained with monoclonal antibodies (mAbs) against different MyHC isoforms: A4.840 (strong affinity for MyHCI;
For fiber size measurement and capillary visualization, mAb 5H2 against laminin α2 chain in muscle fiber basement membrane (Nova Castra Lab, Newcastle, UK) and mAb 4C7 against laminin α5 in capillary basement membrane (Chemicon, Temecula, Calif., USA) were used
The IHC staining process is the same as described earlier
Randomly chosen areas from each section were scanned using a light microscope (Leica DM6000B, Leica Microsystems CMS GmbH, Wetzlar, Germany) equipped with a high-speed fluorescence digital CCD camera (Leica DFC360 FX) connected to an image analysis system (Leica, QWin plus). For each muscle sample, more than 50 fibres (mean 227) were individually analysed in order to obtain a robust morphometric analysis
Based on the staining pattern for the different MyHC mAbs, the fibers were classified as fibers containing solely MyHCI, MyHCIIa or MyHCIIx, or as hybrid fibers co-expressing two MyHC isoforms: MyHCI+IIa or MyHCIIa+IIx. Detailed description of fiber type classification has been described in our earlier study
Estimation of fiber area and number of capillary has been described in detail in a publication from our laboratory
Nuclei in each fiber (NIF) were calculated as all nuclei within each muscle fiber. The number of nuclei in each fiber relative to fiber area (NIFA) was calculated as: NIF/(cross sectional area for each fiber) × 103. This variable measures the nuclear domain in each fiber. Analysis of internal nuclei in each fiber (INIF) was calculated as all the nuclei within each fiber, but without contact to the cell membrane outline by staining for laminin α5.
Normal distribution of data was tested using the Shapiro-Wilk's test and visually inspected through normal quantile plot. Student's unpaired t-test was used to compare measurements between the two groups, but when data was significantly skewed (p<0.05), then the Wilcoxon signed-rank two-sample test with normal approximation was applied. Accordingly, mean and standard deviation (SD) or median and range were used for descriptive statistics. Correlation analysis between AAS dosage and other variables was performed using Pearson correlation and linear regression, and skew data was log-transformed. Orthogonal projections of latent structures discriminant analysis (OPLS-DA) models were used to separate groups (Clean from Doped). One predictive component was calculated (Y), where R2Y display the cumulative percent of the modelled variation in Y, using the X model. Q2Y values display the cumulative percent of the variation in Y that can be predicted by the model according to cross validation (leave one out methods and seven groups), using the X model. The variation modelled of X, using all predictive components and orthogonal components in X, R2X (cum) is a measure of fit, i.e. how well the model fits the X data. JMP 11 (SAS Institute Inc., USA) and SIMCA 13.0 (Umetrics AB, Sweden) were used for all statistical calculations.
Group values of maximal muscle strength and anthropometry were presented in
Variable | Doped | Clean | p |
N = 10 | N = 7 | ||
Body weight (kg) | 108±17 | 110±13 | 0.85 |
Lean body mass (kg) | 89.8±8.2 | 74.6±6.8 | 0.06 |
Lean leg mass (kg) | 28.6±2.5 | 25.5±1.4 | 0.01 |
Personal Bench record (kg) |
205 (155–320) | 190 (145–230) | 0.79 |
Personal Squat record | 254±11 | 265±35 | 0.53 |
Personal Deadlift record (kg) |
257 (150–300) | 269 (245–300) | 0.86 |
Maximal Squat force (N) | 2416±633 | 3302±274 | 0.004 |
Maximal Squat force/Lean body mass (N kg−1) | 29.5±4.0 | 49.8±5.8 | 0.001 |
Maximal Squat force/Lean leg mass (N kg−1) | 88±17 | 130±14 | <0.001 |
Maximal Squat force/Mean fiber area (N μm−2) | 0.33±0.09 | 0.50±0.05 | 0.001 |
Maximal Squat force/Type I fiber area (N μm−2) | 0.38±0.12 | 0.64±0.06 | <0.001 |
Maximal Squat force/Type IIa fiber Area (N μm−2) | 0.28±0.08 | 0.40±0.06 | 0.009 |
Fiber area (μm2) |
7744 (4731–16330) | 6733 (5668–8567) | 0.70 |
Type I fiber area (μm2) |
6511 (3734–15208) | 5189 (4408–6139) | 0.30 |
Type IIa fiber area (μm2) |
9066 (4820–17446) | 8489 (7144–11448) | 0.78 |
Capillary density (n μm−2); CD | 218±43 | 182±41 | 0.12 |
Capillaries/Fiber (n); CAF | 3.93±0.70 | 3.05±0.42 | 0.02 |
Capillaries/Type I fiber (n); CAFI | 4.24±0.60 | 3.16±0.49 | 0.003 |
Capillaries/Type IIa fiber (n); CAFIIa | 4.08±0.66 | 2.94±0.37 | 0.002 |
Capillaries/Mean fiber area (n μm−2); CAFA | 0.55±0.12 | 0.46±0.11 | 0.20 |
Capillaries/Type I fiber area (n μm−2); CAFAI | 0.69±0.16 | 0.62±0.11 | 0.33 |
Capillaries/Type IIa fiber area (n μm−2); CAFAIIa | 0.45±0.10 | 0.36±0.09 | 0.09 |
Nuclei/Type I fiber (n); NIFI | 2.20±0.11 | 1.83±0.13 | 0.04 |
Nuclei/Type IIa fiber (n) |
3.84 (2.5–6.0) | 3.34 (2.6–4.1) | 0.25 |
Nuclei/Type I fiber area (n μm−2) × 1000; NIFAI | 0.37±0.10 | 0.36±0.08 | 0.83 |
Nuclei/Type IIa fiber area (n μm−2) × 1000; NIFAIIa | 0.46±0.10 | 0.40±0.06 | 0.21 |
Internal nuclei/Fiber (n) |
0.07 (0.01–0.25) | 0.07 (0.01–0.36) | 0.98 |
Wilcoxon signed rank test [median (min-max)].
Group values of measurements were presented in
No significant difference in mean fiber area of either type I or type IIa was observed between the Doped and the lean athletes. However, the doped athletes presented 15% larger in mean fibre area and large variation in fibre area compared to the clean athletes (
Sections from one Doped athlete using higher (A; >2500 mg·week−1) and one using lower doses AAS (B;<500 mg·week−1), and from one Clean athlete (C). Doped athletes with higher doses AAS showed larger fiber areas (A) than Doped athletes with lower AAS doses (B) and Clean athletes (C). More capillaries and nuclei around each type I fiber were observed in the Doped athletes (A and B) compared to Clean (C). Internal nuclei are marked with arrows in A.
When capillary measurements were expressed as CD, no difference between the Doped and the Clean groups was observed; however, when expressed as CAF, the Doped group had significantly higher values in both type I and type IIa fibers. When CAF was compensated for fiber area (CAFA), the significant difference between the two groups disappeared for both fiber types (
More nuclei per fiber (NIF) were observed in type I fiber of the Doped group compared to the clean group (
The average number of INIF (internal nuclei/fiber) was very low and did not show statistical difference between the two groups.
Because most blood hormone concentration were not normally distributed, data was analysed by non-parametric statistics (Wilcoxon signed rank, Chi2 approximation), and presented as median and minimum - maximum (
Lutenizing hormone (LH; E • L−1) | Pituitary gland | 1.2–9.6 | 2.5 (1.2–4.8) | 0 (0–1.2) | <0.001 |
Folicular stimulating hormone (FSH; E • L−1) | Pituitary gland | 1.0–12.5 | 3.0 (1.0–4.2) | 0 (0–2.3) | <0.001 |
17-alfa-Hydroxiprogesteron (17-OH-prog; nmol • L−1) | Adrenal glands | < 10 | 1.85 (0.70–7.00) | 0.30 (0.30–0.70) | 0.001 |
Alanine aminotransferas ALAT (μkat • L−1) | Liver | < 1.2 | 0.55 (0.48–0.70) | 0.77 (0.59–2.08) | 0.005 |
Aspartate aminotransferas ASAT (μkat • L−1) | Liver | < 0.76 | 0.47 (0.33–0.65) | 0.68 (0.39–1.92) | 0.01 |
Prolactin (Prol; μg • L−1) | Pituitary gland | 3–13 | 6 (4–13) | 10 (6–16) | 0.01 |
Urea (mmol L−1) | Muscle/Kidney | 3.2–8.1 | 7.4 (4.5–9.6) | 4.8 (2.3–5.7) | 0.02 |
Creatin Kinase (CK; μkat • L−1) | Muscle | 0.8–6.7 | 4.8 (3.9–11.8) | 11.8 (4.2–85.2) | 0.03 |
Testosterone (Testo; nmol • L−1) | Androgen | 6.3–16 | 11 (7.1–18) | 35 (3.8–130) | 0.08 |
Pro-brain natriuretic peptide (ProBNP; ng • L−1) | Heart | < 84 | 14 (9–31) | 10 (0–32) | 0.09 |
Androstendione (nmol • L−1) | Adrenal glands | 3.2–9.9 | 2.0 (1.2–2.8) | 3.1 (0.6–14) | 0.09 |
Creatinin (Crea; μmol • L−1) | Muscle/Kidney | < 100 | 94 (84–133) | 88 (77–112) | 0.12 |
Apolipoprotein B (ApoB, g • L−1) | Liver/Intestine | 0.50–1.50 | 0.87 (0.76–1.44) | 1.24 (0.83–1.61) | 0.16 |
Sexual hormone binding globuline (SHBG; nmol • L−1) | Liver | 15–56 | 29 (18–46) | 23 (3.9–54) | 0.19 |
Apolipoprotein A (ApoA; g • L−1) | Liver/Intestine | 1.10–1,80 | 1.19 (1.02–1.60) | 1.09 (0–1.36) | 0.20 |
Troponin I (Trop I; μg • L−1) | Heart | < 0.03 | 0 (0–0.11) | 0 (0–0) | 0.20 |
Estradiol (E2K; pmol • L−1) | Estrogen | 50–150 | 81 (55–102) | 182 (25–425) | 0.25 |
Growth hormone (GH; μg • L−1) | Hypophysis | none | 0.10 (0.01–4.80) | 0.05 (0–3.1) | 0.33 |
Insulin like growth factor IGF-I (μg • L−1) | Anabolism | 120–420 | 194 (95–384) | 158 (114–259) | 0.35 |
Dehydroepiandrosterone sulfate (DHEAS; μmol • L−1) | Adrenal glands | 2.4–13 | 8.4 (6.0–11) | 7.8 (1.3–12) | 0.36 |
Cystatin C (CystC; mg • L−1) | Kidney | < 0.99 | 0.89 (0.70–0.96) | 0.84 (0.75–1.00) | 0.59 |
High sensitive C –reactive protein (HCRP; mg • L−1) | Heart/Inflammation | < 3 | 0.4 (0.3–5–4) | 1.37 (0–25.6) | 0.73 |
Lipoprotein (a) (Lp(a); mg • L−1) | Heart | < 700 | 356 (84–1463) | 372 (222–782) | 0.81 |
Thyroid stimulating hormone (TSH; mE • L−1) | Metabolism | 0.4–4.7 | 1.5 (0.6–2.0) | 1.7 (0.8–2.8) | 0.85 |
Cortisol (Cortis; nmol • L−1) | Adrenal glands | 100–800 | 415 (281–550) | 402 (146–498) | 0.87 |
Alkaline phosphatase (ALP; μkat • L−1) | Liver | < 1.20 | 1.0 (0.7–1.6) | 1.15 (0.70–1.70) | 0.88 |
Wilcoxon signed rank test [median (min-max)]. * From the Karolinska University Laboratory (
Correlations between AAS weekly intake and muscle performance: A) personal record (kg; R2 = −0.16, p = 0.86) and B) maximal squat force (N; R2 = 0.34, p = 0.06), and models for the effects of AAS intake on relative muscle performance: C) maximal squat force per lean leg mass (N·g−1; R2 = 0.02, p = 0.32) and D) maximal squat force per fiber area (N·μm−1; R2 = −0.14, p = 0.98). The residual of subject G is outlier (p = 0.003, Shapiro-Wilk W test) and when removed, the regression is significant between force per fiber area and AAS intake (N·μm−1; R2 = 0.57, p = 0.02).
r | p | r | p | |
Anthropometry | ||||
Body Weight (kg) | 0.33 | 0.35 | 0.21 | 0.59 |
Lean body mass (kg) | 0.62 | 0.05 | 0.43 | 0.25 |
Lean leg mass (kg) | 0.53 | 0.12 | 0.27 | 0.48 |
Performance | ||||
Maximal Squat force (N) | 0.65 | 0.06 | 0.75 | 0.03 |
Maximal Squat force/Lean body mass (N • g−1) | 0.51 | 0.17 | 0.83 | 0.01 |
Maximal Squat force/Lean leg mass (N • g−1) | 0.55 | 0.13 | 0.88 | 0.004 |
Maximal Squat force/Mean fiber area (N • μm−2) | −0.01 | 0.98 | 0.76 | 0.01 |
Maximal Squat force/Type I fiber area (N • μm−2) | −0.01 | 0.98 | 0.80 | 0.02 |
Maximal Squat force/Type IIa fiber Area (N • μm−2) | −0.08 | 0.85 | 0.60 | 0.12 |
Morphology | ||||
Mean fiber area (μm2)Log10 | 0.69 | 0.02 | 0.24 | 0.53 |
Type I fiber area (μm2)Log10 | 0.70 | 0.06 | 0.25 | 0.51 |
Type IIa fiber area (μm2)Log10 | 0.62 | 0.06 | 0.24 | 0.53 |
Capillary density (μm2); CD | −0.33 | 0.35 | 0.52 | 0.15 |
Capillaries/Fiber (n); CAF | 0.47 | 0.17 | 0.59 | 0.10 |
Capillaries/Type I fiber (n); CAFI | 0.64 | 0.05 | 0.61 | 0.08 |
Capillaries/Type IIa fiber (n); CAFIIa | 0.29 | 0.41 | 0.49 | 0.18 |
Capillaries/Mean fiber area (n • μm−2); CAFA | −0.39 | 0.27 | 0.57 | 0.11 |
Capillaries/Type I fiber area (n • μm−2); CAFAI | −0.43 | 0.22 | 0.47 | 0.20 |
Capillaries/Type IIa fiber area (n • μm−2); CAFAIIa | −0.45 | 0.2 | 0.45 | 0.23 |
Nuclei/Fiber (n); NIF | 0.30 | 0.40 | 0.21 | 0.59 |
Nuclei/Type I fiber (n); NIFI | 0.26 | 0.47 | 0.14 | 0.73 |
Nuclei/Type IIa fiber (n)LOG; NIFIIa | 0.33 | 0.36 | 0.28 | 0.46 |
Nuclei/Fiber area (n • μm−2); NIFA | −0.63 | 0.05 | −0.11 | 0.78 |
Nuclei/Type I fiber area (n μm−2); NIFAI | −0.64 | 0.05 | −0.16 | 0.69 |
Nuclei/Type IIa fiber area (n • μm−2); NIFAIIa | −0.61 | 0.06 | −0.09 | 0.81 |
Inner Nuclei/Fiber (n); INIF | 0.55 | 0.10 | 0.12 | 0.76 |
Log10; transformed for normal distribution before calculations.
In OPLS-DA analysis, a score scatter plot (
Morphological and performance variables (N = 8) are used in an OPLS-DA model to separate Doped (N = 9) from Clean (N = 6) subjects. Regression = 0.93, and prediction by cross-validation = 0.92, p<0.0001, Fisher's exact probability test. All nine Doped subjects and six of seven Clean are correctly classified, leaving one Clean un-classified. Variables of importance are displayed in
From the OPLS-DA model in
The main findings of the study were that the doped athletes had higher lean mass, capillary density and myonuclei density, but lower maximal squat force relative to muscle mass and to fiber area, compared to the clean athletes. The Doped group also had a tendency towards larger fibers, although not significant, most likely due to large variations in fibre area. Low levels of LH and FSH, and high levels of prolactin in some individuals indicate a disturbed pituitary gland function with possible negative effects on reproductive function. High levels of ALAT, ASAT and CK in some individuals suggest that long term use of AAS could damage both liver and muscle tissue. However, no correlations between AAS intake and hormone levels was observed. Thus, testosterone levels at time of sampling cannot explain alternations in these variables, rather concentrations outside clinical limits must stem from long-term supplementation of AAS. Multivariate statistics showed that a combination of eight morphological parameters could clearly separate the doped from the clean athletes. Correlation analysis revealed significant positive correlations between AAS dosage and relative muscle force. The results support previous findings that AAS administration could induce enhancement in both muscle mass and muscle strength, and that the improvements are AAS dose-dependent
Despite abundant studies on the effects of AAS on skeletal muscle, many results are contradictory
Increased muscle mass in subjects using AAS has been proposed to result from muscle hypertrophy alone
Not many studies have examined the effects of AAS on muscle capillaries. In a previous study of 20 weeks of graded testosterone enanthate injection (25, 50, 125, 300, or 600 mg), Sinha-Hikim et al.
It has been shown that combined administration of androgens and resistance training is associated with greater gains in lean body mass, muscle size, and maximal voluntary strength than either intervention alone
It is worth to notice that compared to the Clean group, the Doped group presented larger variations in many of the measurements like leg lean mass (Doped, 24.6–32.6 kg vs. Clean, 22.8–26.9 kg), leg maximal strength (Doped, 1823–3242 N vs. Clean, 2799–3570 N) and muscle fibre size (Doped, 6055–16330 µm2 vs. Clean, 5668–8567 µm2). Giorgi et al.
Previous studies have shown that testosterone administration was associated with a dose-dependent increase in skeletal muscle mass, leg strength and power
In line with laboratory intervention studies
While all the Doped athletes have used AAS, the mix and quality of the substance are unknown. This may confound the estimation of AAS dosage as well as the effects on muscle morphology and performance. Additionally, post-study subjects de-coding revealed that Doped group was older and composed of athletes involved in bodybuilding and strongmen events, while Clean athletes were all power-lifters. Consequently, training regiments were slightly different, even though all aiming at increasing muscle strength. Consequences for interpretation of data are several: 1) Doping controls implemented for power-lifters in Sweden has reduced the number of doped athletes, while the same anti-doping efforts have not been taken in other power events. 2) Because the higher age in the Doped group, it can be speculated that athletes in their later career are more prone to AAS. 3) The dose-response effects of AAS on muscle morphology and performance were in agreement with previous intervention studies.
To further explore the effects of long term AAS supplementation on skeletal muscles, more advanced techniques, such as proteomics and metabolomics should be applied in tissue analysing. Again, we have to emphasize that the current study design is hard to be replicated in laboratory due to the extreme doses and duration of AAS supplementation.
We thank M Enerstedt, A-K Olofsson and P Boman for excellent technical assistance. Associate Professor, MD Stefan Arver is acknowledged for consultation when selecting blood variables for analysis.