While recent studies have begun to address how hormones mediate whole-animal performance traits, the field conspicuously lags behind research conducted on humans. Recent studies of human steroid use have revealed that steroid use increases muscle cross-sectional area and mass, largely due to increases in protein synthesis, and muscle fiber hypertrophy attributable to an increased number of satellite cells and myonuclei per unit area. These biochemical and cellular effects on skeletal muscle morphology translate into increased power and work during weight-lifting and enhanced performance in burst, sprinting activities. However, there are no unequivocal data that human steroid use enhances endurance performance or muscle fatigability or recovery. The effects of steroids on human morphology and performance are in general consistent with results found for nonhuman animals, though there are notable discrepancies. However, some of the discrepancies may be due to a paucity of comparative data on how testosterone affects muscle physiology and subsequent performance across different regions of the body and across vertebrate taxa. Therefore, we advocate more research on the basic relationships among hormones, morphology, and performance. Based on results from human studies, we recommend that integrative biologists interested in studying hormone regulation of performance should take into account training, timing of administration, and dosage administered when designing experiments or field studies. We also argue that more information is needed on the long-term effects of hormone manipulation on performance and fitness.
One of the most widely discussed and controversial arenas of human performance concerns the use of steroid supplements to enhance athletic ability for a variety of sports, ranging from bicycling to baseball. There is strong evidence that human athletes have attempted to enhance their athletic performance using steroids since the 1950s, but whether, and in which sports, steroids are actually effective remains controversial (reviewed by Ryan 1981; George 2003; Hartgens and Kuipers 2004). In general, steroids used by athletes encompass a wide variety of forms of the androgen testosterone (George 2003), and most seem to have the classical androgenic and anabolic effects on men, although steroid use by women cannot be ignored (Malarkey et al. 1991; Gruber and Pope 2000). Alternative forms of testosterone (e.g., testosterone enanthate, methandrostenolone) are typically used by those desiring enhanced performance because ingested or injected testosterone is quickly metabolized into inactive forms (Wilson 1988). Thus, studies of humans that we cite involve testosterone derivatives. Early studies of the effects of steroids on human performance, however, had major flaws in design, such as lack of control groups and a double-blind procedure, the presence of confounding factors (e.g., differences in level of exercise and in motivation), and inappropriate statistical techniques (reviewed by Bhasin et al. 2001; George 2003). These problems left open for many years the question of whether, and in what capacity, steroids actually enhance athletic performance, until more recent studies conclusively showed significant effects of steroids.
The topic of steroid effects on human athletic performance is germane to an emerging field of research investigating hormonal effects on animals’ performance (e.g., sprint speed, endurance capacity, bite-force capacity) (Husak et al. 2009a), as testosterone may exert general effects on performance across widely divergent vertebrate taxa. Our goal in this review is to interpret the effects of steroids on human performance in this broader context of hormonal effects across a wider range of taxa. We are particularly interested in drawing lessons and potential avenues of research for animal biologists from published research on humans. We have performed a selective review of studies examining how humans' use of steroids affects skeletal muscle physiology and subsequent athletic performance. While studies of performance on nonhumans have dealt extensively with the effects of morphological traits on performance and the impact of performance on individual fitness (Arnold 1983; Garland and Losos 1994; Irschick and Garland 2001; Irschick et al. 2007, 2008; Husak et al. 2009a), there has been relatively little synthetic discussion of how hormones affect performance in non-human animals. We also point the reader towards several recent reviews of steroid use and performance by humans for details not discussed in our review (Bhasin et al. 2001; George 2003; Hartgens and Kuipers 2004).
General effects of testosterone on the phenotype of males
The development of primary and secondary sexual characteristics is stimulated by testosterone in vertebrate males, and these effects can be either organizational or activational in nature (Norris 1997; Hadley 2000). Organizational effects tend to occur early in development, and during a critical window of time, thereby resulting in permanent effects. On the other hand, activational effects occur in adults, and the effects are typically temporary (Arnold and Breedlove 1985). The hypothalamus stimulates production of gonadotropin-releasing hormone, which in turn stimulates production of luetenizing hormone in the anterior pituitary. Luetenizing hormone then stimulates production of testosterone in the Leydig cells of the testes. Testosterone then circulates throughout the body where it exerts effects on multiple target tissues that have the appropriate receptors or appropriate enzymes (e.g., aromatase or 5α-reductase) to convert testosterone for binding to other types of receptors (Kicman 2008). The widespread effects of circulating levels of testosterone on aggression, secondary sexual traits, and growth of skeletal muscle in males of many vertebrate species are well-documented (Marler and Moore 1988; Wingfield et al. 1990; Ketterson and Nolan 1999; Sinervo et al. 2000; Ketterson et al. 2001; Oliveira 2004; Adkins-Regan 2005; Hau 2007; contributions in this issue). In particular, production of testosterone by males has been linked with the expression of color and behavioral display signals, as well as aggression (Marler and Moore 1988; Kimball and Ligon 1999; Hews and Quinn 2003; Adkins-Regan 2005; Cox et al. 2008) and increased growth (Fennell and Scanes 1992; Borski et al. 1996; Cox and John-Alder 2005), although this latter effect may depend on specific selective pressures on males (Cox and John-Alder 2005).
Effects of testosterone on the physiology of human skeletal muscle
Testosterone has multiple effects on skeletal muscle at the biochemical and cellular levels, but the direct cause-and-effect relationships among these effects are still unclear (Sinha-Hikim 2002; Hartgens and Kuipers 2004). The studies that we discuss here, and throughout the paper are from experiments or correlative studies conducted on adult individuals such that the effects seen are activational in nature, causing rather rapid changes to the phenotype. Increased testosterone causes increased protein synthesis by muscle cells (Griggs et al. 1989; Kadi et al. 1999; Hartgens and Kuipers 2004), which is necessary for anabolic effects and an increase in lean muscle mass. Sinha-Hikim et al. (2002) found a dose-dependent increase in the mean number of myonuclei found in skeletal muscle fibers (vastus lateralis muscle) with testosterone supplementation, as well as in the number of myonuclei per fiber (see also Eriksson et al. 2005). This increase was also associated with an increase in the number of satellite cells in the muscle tissue (but see Eriksson et al. 2005). Satellite cells are progenitor cells found external to muscle fibers that are incorporated into fibers and promote repair and growth of the muscle (Kadi and Thornell 2000; Reimann et al. 2000). However, the mechanism by which testosterone causes an increase in the number of satellite cells is unknown and could be due to testosterone (1) promoting cell division of satellite cells, (2) inhibiting apoptosis of satellite cells, or (3) causing differentiation of stem cells into satellite cells (Sinha-Hikim 2002). In any case, the functional implications for these findings are clear. More satellite cells likely result in more myonuclei per fiber, which, combined with increased protein synthesis, contribute to increases in muscle growth via an increased number and hypertrophy of muscle fibers (Kadi 2000; Kadi and Thornell 2000).
Testosterone also appears to cause a dose-dependent increase in the cross-sectional area of muscle fibers, although details about which types of fibers are affected and where in the body this occurs remains equivocal. Testosterone may increase the cross-sectional area of both type I (oxidative “slow twitch”) and type II (glycolytic “fast twitch”) fibers simultaneously after administration (Sinha-Hikim 2002; Eriksson et al. 2005), but other studies have shown greater increases in type I than in type II fibers (Hartgens et al. 1996; Kadi et al. 1999; also in growing rats, Ustunelet al. 2003), increased size in only type I fibers (Alén et al. 1984; Kuipers et al. 1991, 1993), or increased size in only type II fibers (Hartgens et al. 2002). These mixed results are intriguing, because they suggest that different parts of the body, and, hence, different performance traits, may be affected differently by elevated testosterone levels. The likely mechanism for these differences is variation in density of receptors within the myonuclei of muscle fibers in different regions of the body (Kadi 2000; Kadi et al. 2000). An alternative hypothesis is that different types of fiber have differing relationships between the number of internal myonuclei and muscle cross-sectional area during hypertrophy (Bruusgaard et al. 2003). That is, some types of fibers may have internal myonuclei that can serve larger “nuclear domains” than can other types of fibers (reviewed by Gundersen and Bruusgaard 2008). If either of these hypothesized mechanisms is correct, then circulating levels of testosterone may only explain a portion of inter-individual (or interspecific) variation in performance. Testosterone may also stimulate changes in the proportions of types of fibers in muscles (Holmang et al. 1990; Pette and Staron 1997), although evidence for this effect in humans is mixed. For example, Sinha-Hikim et al. (2002) did not observe a change in the proportions of type I and type II fibers after administration of testosterone.
Changes in lower-level traits (e.g., protein synthesis, number of satellite cells, cross sectional area of muscle fibers) after testosterone supplementation, as described above, thus, result in changes at the whole-muscle level and explain many of the classic effects of testosterone that are desired by humans using steroids. That is, increasing testosterone via steroid use increases body weight, lean body mass, as well as cross-sectional area, circumference, and mass of individual muscles (i.e., “body dimensions”); however, there are numerous studies with contradictory results, finding no change in one, or all, of these traits, depending on the drug used, the dose taken, and the duration of use (reviewed by Bhasin et al. 2001; Hartgens and Kuipers 2004). The finding that testosterone can change muscle physiology and increase whole-muscle size and/or body mass is consistent with results in nonhuman animals. For example, testosterone implants increased size and number of fibers in the sonic muscles of male plainfin midshipman fish (Porichthys notatus) (Brantley et al. 1993). Similarly, testosterone supplementation increased muscle mass and changed contractile properties of trunk muscles of male grey treefrogs (Hyla chrysoscelis) (Girgenrath and Marsh 2003) and of forelimb muscles of male frogs (Xenopus laevis, Regnier and Herrera 1993; Rana pipiens, Sidor and Blackburn 1998).
Effects of testosterone on humans’ performance
Whether steroids actually enhance performance of athletes was a subject of great controversy throughout the 1980s and 1990s (Ryan 1981; Haupt and Rovere 1984; Cowart 1987; Wilson 1988; Elashoff et al. 1991; Strauss and Yesalis 1991; Hartgens and Kuipers 2004), largely due to flaws in design of early studies (see above). However, the past decade has seen a surge in more carefully designed studies that have convincingly tested whether, all else equal, steroids increase performance. Hartgens and Kuipers (2004) found that 21 out of 29 studies they reviewed found an increase in humans’ strength after steroid use, with improvements in strength ranging from 5% to 20%. Storer et al. (2003) found that testosterone caused a dose-dependent increase in maximal voluntary strength of the leg (i.e., amount of weight lifted in a leg press), as well as in leg power (i.e., the rate of force generation). They further tested whether increased muscle strength was due simply to increased muscle mass or to changes in the contractile quality of muscle affected by testosterone, but they found no change in specific tension, or in the amount of force generated per unit volume of muscle. This latter result suggests that, at least for leg-press performance, testosterone increases strength by increasing muscle mass and not by changing contractile properties. Rogerson et al. (2007) found that supraphysiological doses of testosterone increased maximal voluntary strength during bench presses (see also Giorgi et al. 1999) and increased output of work and output of power during cycle sprinting compared to placebo control subjects. Thus, “burst” or “sprint” performance traits appear to be enhanced by increased testosterone, and this is in general agreement with studies of nonhuman animals (John-Alder et al. 1996, 1997; Klukowski et al. 1998; Husak et al. 2007). For example, experimentally elevated levels of testosterone caused increased sprint speed, relative to sham-implanted individuals, in northern fence lizards (Sceloporus undulatus) (Klukowski et al. 1998). These findings contrast with results for endurance events, in which no increase in performance has been detected experimentally in humans (reviewed in George 2003; Hartgens and Kuipers 2004). The finding that endurance by humans is not enhanced by testosterone is unexpected since testosterone may increase hemoglobin concentrations and hematocrit (Alén 1985, but see Hartgens and Kuipers 2004) and exogenous testosterone increases endurance in rats (Tamaki et al. 2001) and male side-blotched lizards (Uta stansburiana) (Sinervo et al. 2000). More studies of the effects of increased testosterone on endurance would help to better clarify these seemingly paradoxical findings. One possibility that might explain species’ differences in endurance is the relative proportion of type I fibers available for enhancement, which likely varies across species (Bonine et al. 2005), although this hypothesis needs explicit testing. Steroid use does not seem to consistently enhance recovery time after strenuous exercise (reviewed in Hartgens and Kuipers 2004), although it may in non-human animals (Tamaki et al. 2001). Storer et al. (2003) also found no change in fatigability (i.e., the ability of a muscle to persist in performing a task) of muscle during exercise, which is consistent with other studies (George 2003).
One of the problems in early studies of steroid effects was that the participants’ history of training and exercise while taking steroids was not taken into account or controlled (Bhasin et al. 2001; George 2003; Hartgens and Kuipers 2004). Recent studies have shown that the presence or absence of exercise training during testosterone supplementation can have a marked impact on how much performance is enhanced, thus complicating results when training is not controlled. Bhasin et al. (2001) reviewed several examples of such results. They pointed out that testosterone supplementation alone may increase strength from baseline levels, but so will exercise alone with a placebo, such that strength levels with exercise alone are comparable to those with testosterone addition alone (Bhasin et al. 1996). Testosterone supplementation while undergoing exercise training typically has the greatest increase in strength compared to exercise only or testosterone only (Bhasin et al. 1996, 2001). These findings are consistent with those of others (reviewed by George 2003). Indeed, George (2003) suggested that steroids will only consistently enhance strength if three conditions are met: (1) steroids are given to individuals who have been training and who continue to train while taking steroids, (2) the experimental subjects have a high protein diet throughout the experiment, and (3) changes in performance are measured by the technique with which the individuals were training while taking steroids. That is, one may, or may not, find a change in bench-press performance if individuals trained with leg presses, and not bench presses, while taking steroids. We note that the confounding effect of training is a rather intuitive finding, but it does point out potential problems in studies of non-human animals, specifically laboratory studies, which we address below.
Implications of studies of humans for studies of nonhuman animals
STEROIDS is an international research journal devoted to studies on all chemical and biological aspects of steroidal moieties. The journal focuses on both experimental and theoretical studies on the biology, chemistry, biosynthesis, metabolism, molecular biology, physiology and pharmacology of steroids and other molecules that target or regulate steroid receptors. Manuscripts presenting clinical research related to steroids, steroid drug development, comparative endocrinology of steroid hormones, investigations on the mechanism of steroid action and steroid chemistry are all appropriate for submission for peer review. STEROIDS publishes both original research and timely reviews. For details concerning the preparation of manuscripts see Instructions to Authors, which is published in each issue of the journal.
Please bookmark this URL: http://www.elsevier.com/locate/steroids
Benefits to authors
We also provide many author benefits, such as free PDFs, a liberal copyright policy, special discounts on Elsevier publications and much more. Please click here for more information on our author services.
Please see our Guide for Authors for information on article submission. If you require any further information or help, please visit our Support CenterHide full Aims & Scope