Test addicitve? new research in animals.

--------------------------------------------------------------------------------

There's actually some interesting info in this longish article.

Reinforcing aspects of androgens

Ruth I. Wood,

Are androgens reinforcing? Androgenic?anabolic steroids (AAS) are drugs of abuse. They are taken in large quantities by athletes and others to increase performance, often with negative long-term health consequences. As a result, in 1991, testosterone was declared a controlled substance. Recently, Brower [K.J. Brower, Anabolic steroid abuse and dependence. Curr. Psychiatry Rep. 4 (2002) 377?387.] proposed a two-stage model of AAS dependence. Users initiate steroid use for their anabolic effects on muscle growth. With continued exposure, dependence on the psychoactive effects of AAS develops. However, it is difficult in humans to separate direct psychoactive effects of AAS from the user's psychological dependence on the anabolic effects of AAS. Thus, studies in laboratory animals are useful to explore androgen reinforcement. Testosterone induces a conditioned place preference in rats and mice, and is voluntarily consumed through oral, intravenous, and intracerebroventricular self-administration in hamsters. Active, gonad-intact male and female hamsters will deliver 1 ?g/?l testosterone into the lateral ventricles. Indeed, some individuals self-administer testosterone intracerebroventricularly to the point of death. Male rats develop a conditioned place preference to testosterone injections into the nucleus accumbens, an effect blocked by dopamine receptor antagonists. These data suggest that androgen reinforcement is mediated by the brain. Moreover, testosterone appears to act through the mesolimbic dopamine system, a common substrate for drugs of abuse. Nonetheless, androgen reinforcement is not comparable to that of cocaine or heroin. Instead, testosterone resembles other mild reinforcers, such as caffeine, nicotine, or benzodiazepines. The potential for androgen addiction remains to be determined.


1. Biological context: why are androgens reinforcing?

In a natural environment, testosterone secretion is intimately related to reinforcing social behaviors, including mating and aggression [1] and [2]. While castrated males do not mate and show little offensive aggression, testosterone replacement restores both sexual and agonistic behaviors. Although testosterone and other gonadal steroids facilitate social behavior, the converse is also true. That is, steroid secretion can be modified by social interactions. Successful expression of mating or aggression stimulates testosterone secretion in gonad-intact males [3], [4], [5], [6] and [7], and defeat depresses testosterone further [5], [6] and [7]. In this regard, dominant males win more agonistic encounters, copulate with greater frequency, and have higher levels of testosterone, compared to subordinates [4], [5], [6] and [7]. Together, the foregoing observations support the hypothesis that testosterone amplifies the reinforcing or aversive value of social behaviors, thereby serving as an aid to discriminating successful and unsuccessful social interactions. This suggests the additional possibility that androgens such as testosterone may be reinforcing, independent of social behavior. This review summarizes recent research on testosterone reinforcement, with particular relevance to the abuse of androgenic?anabolic steroids (AAS).
2. Defining the problem: AAS abuse in humans
2.1. Who is taking steroids?

Through their androgenic, anabolic, and psychotropic actions, androgens enhance performance, both athletic achievement and ?competition? in nonathletic social situations [8]. Not surprisingly, AAS are abused, both by athletes and casual users. Although AAS were banned from Olympic competition in 1975, steroid abuse continues. In unannounced drug tests at national competitions, up to 50% of athletes tested positive for AAS [9]. In response to widespread abuse, testosterone was declared a controlled substance in 1991. Like opiates and psychostimulants, AAS abuse has negative health consequences for users, including endocrine, hepatic, and cardiovascular disturbances [10]. In addition, traffic in illegal steroids has negative societal impact, including ingestion/injection of impurities mixed with steroids, crime from selling and purchasing, and disease (AIDS, hepatitis) from needle sharing [8].

AAS abuse is endemic in power sports (up to 95% of professional football players, 80?99% of male body builders) [8]. However, a common misconception is that AAS abuse is restricted to athletes and bodybuilders. Many noncompetitors also use steroids ?just for their ego,? ?to get dates? to be part of a gang,? and for ?feeling good about yourself and being admired and looked at? [8]. These reports suggest an effect of androgens not only on somatic tissues, but also in the CNS. AAS abuse in nonathletes is particularly common among adolescent boys (7% of HS seniors) [9], for whom exogenous androgens can limit the pubertal growth spurt by promoting epiphyseal closure.
2.2. What are they taking? How do they take it?

AAS users ingest a variety of androgenic compounds orally, transdermally, or by injection, and can achieve circulating androgen concentrations up to 100× the physiologic range for an adult male [11]. An interesting feature of AAS abuse is that users are often well educated about the substances that they use and well connected through muscle magazines and web sites. Thus, information about preferred types of AAS, patterns of use, and side effects is readily available at https://www.anabolics.com, https://www.musclehead.com and similar web sites. Most of these acknowledge effects of AAS on mood and behavior, both during active steroid use and during the ?crash? that follows the end of a steroid cycle [12] and [13].

The majority of AAS is derived from testosterone, and all have both anabolic (muscle building) and androgenic (masculinizing) properties. AAS users aim to maximize their anabolic effects, while limiting androgenic actions. However, there is no ?pure? anabolic steroid. Native testosterone has several disadvantages for steroid supplementation: it has relatively low potency and is rapidly metabolized by the liver [14]. Modifications to the testosterone molecule include hydroxylation at the C10 position to increase the relative potency (e.g., nandrolone), esterification to slow the release rate into circulation (e.g., testosterone cypionate), or alkylation at the C17 position to reduce first-pass metabolism in the liver, thereby permitting oral delivery (e.g., stanozolol). Modifications to testosterone also affect metabolism to estrogen or the more potent androgen, dihydrotestosterone (DHT). Nonaromatizable androgens, which cannot be converted to estrogen (e.g., dianabol, masteron), produce fewer estrogen-dependent side effects (gynecomastia). Nonreducible androgens, which cannot be converted to DHT (e.g., oxandrolone, oxymetholone), often have low potency but may have a better anabolic/androgenic ratio [15].

Most steroid users do not limit themselves to a single dose or a single type of steroid. Instead, users combine oral and injectable steroids (?stacking?) in increasing and decreasing weekly cycles (?pyramiding?) [16]. As an example, in the fourth week of the ?ultimate Stack? [12], the user would take over 2 g of steroids, including 1 g of oxymetholone, a 17?-alkylated androgen with hepatotoxicity.
2.3. Why are they taking steroids? Evidence for androgen reinforcement in humans

Much of the initial motivation for AAS use is undoubtedly secondary to the physical benefits of increased strength and muscle mass. Recently, Brower [17] proposed a two-stage model of AAS dependence. According to the model, anabolic effects of AAS on muscle growth account for the first stage of steroid use. However, with chronic exposure, users can develop physical and psychological dependence on AAS, as defined by DSM-III-R [18]. In this regard, testosterone reinforcement is often hinted at, but few systematic studies have been performed in humans or animals. Anecdotal evidence in humans suggests that AAS abusers take more androgens with each successive drug cycle [19], experience a ?crash? after a drug cycle is complete [20], and work out more intensely with each cycle [8]. There are also reports that AAS may possess euphorigenic effects (Refs. [18] and [21]; see also Refs. [22], [23] and [24]). In the 1980s, East German scientists developed an androgen nasal spray to enhance aggression and athletic competitiveness without systemic effects. One athlete described the result as ?like a volcano? [25]. Similarly, intranasal 4,16-androstadien-3-one induces an amphetamine-like ?high? in human volunteers [26]. However, it is important to note that the addiction potential for testosterone is not comparable to that of highly addictive drugs such as cocaine or opiates: former opiate addicts did not report euphoria following a single intramuscular (im) injection of testosterone [27].
3. Steroids and experimental models of reward: limitations of human studies

While it is clear that AAS are abused, with many of the same negative consequences for health and society as abuse of addictive drugs, defining the potential for AAS addiction in humans has been problematic because it is difficult in humans to separate direct psychoactive effects of AAS from the user's psychological dependence on the anabolic effects of AAS. Thus, studies in laboratory animals are useful to explore androgen reinforcement. Several experimental models have been developed to study reward in laboratory animals [28]. Three important methodologies are conditioned place preference, self-administration, and brain stimulation reward. These have been applied to studies of androgen reinforcement.
3.1. Conditioned place preference

Testosterone reinforcement has been demonstrated using conditioned place preference in male rats [29] and [30] and mice [31]. In this model, the test substance is repeatedly paired with a unique environment (e.g., a particular chamber in the testing apparatus). Once the animal associates the reinforcing test substance with that environment, it will seek out the environment even in the absence of reward. Initial studies [29], [30] and [31] used systemic androgen treatment. However, when delivered as small volumes in an aqueous solution, rats will also form a conditioned place preference to testosterone injections directly into nucleus accumbens (Acb) [32]. This suggests that testosterone reinforcement is mediated by the brain, rather than by peripheral androgen action.

In a related conditioned taste aversion test, estrogen decreased, but testosterone increased, the consumption of a saccharin solution [33]. With this model, rats learn to associate the taste of saccharin with a steroid injection. A reduction in saccharin intake after estrogen treatment signifies that estrogen is aversive. By contrast, the increase in saccharin consumption in testosterone-treated males suggests that testosterone is reinforcing.
3.2. Brain stimulation reward

There has been only scant research on testosterone and brain stimulation reward. With this model, lever pressing for direct intracranial electrical stimulation is compared in drug-treated and drug-free states [28]. Most addictive drugs will stimulate lever pressing, which is thought to reflect a summation of drug- and stimulation-induced reward. Estradiol plus progesterone potentiates brain stimulation reward in female rats [34]. Testosterone does not directly stimulate lever pressing in male rats, but it does enhance responding in the presence of amphetamine [35]. More research works on the ability of AAS to influence lever pressing for brain stimulation reward must be done before any conclusions can be made.
3.3. Self-administration

As a model of addiction, self-administration has the greatest face validity because the animal controls drug delivery [28]. There is precedent for self-administration of steroid hormones: male rats self-administer corticosterone orally [36] and intravenously [37]. A key advantage of self-administration is the potential for chronic androgen exposure. In humans, testosterone replacement in hypogonadal men is achieved by im injection or transdermal patch [20] to produce a relatively constant level of androgen in circulation. AAS abuse is typically achieved orally or by im injection [10], and users take steroids continually for weeks or months. Other models for testing reward, such as conditioned place preference, do not allow for long-term testosterone delivery. Our studies have used oral, intravenous (iv), and intracerebroventricular (icv) testosterone self-administration as a model to test chronic androgen reinforcement.
3.4. Oral testosterone self-administration

We assessed oral testosterone self-administration in male hamsters using both two-bottle choice and food-induced drinking [38]. In the choice test, the animal has the opportunity to freely consume solutions (testosterone or vehicle) from two bottles in his home cage. In our initial test, males preferred an aqueous solution of 200 mg/ml testosterone over vehicle (3% ethanol). When the taste of testosterone was masked by unsweetened Kool-Aid added to both solutions, self-administration of 400 ?g/ml testosterone developed slowly over 2 weeks. Unlike testosterone, cholesterol was not preferred in a two-bottle choice test, suggesting that reinforcement is not a general property of all sterols. By contrast, males expressed a strong preference for 0.1% saccharin within the first day of exposure. The implication is that oral testosterone is reinforcing, but this effect is not explained by a highly preferred taste.

With testosterone or other slow-acting stimuli in a two-bottle choice test, it can be difficult for the animal to associate the reward with a specific solution. Therefore, in subsequent studies, we used a model of food-induced drinking that takes advantage of the observation that animals drink more when food is present (Fig. 1). The method was developed to study ethanol self-administration. When a testosterone solution (0.4?3 mg/ml) was offered with food for 3 h/day, hamsters drank readily (3.0±0.2 ml). When food was withdrawn, fluid intake was maintained (3.1±0.3 ml). By contrast, water consumption was minimal when food was not present (1.1±0.1 ml). This demonstrates that hamsters will voluntarily consume testosterone under conditions that do not support substantial water intake. Together, our studies with oral testosterone consumption provide the first evidence for testosterone self-administration in laboratory animals.


3.5. Intravenous testosterone self-administration

While oral self-administration in rodents has parallels with oral AAS intake in humans, there are behavioral and physiologic limitations to this method. In particular, we cannot eliminate potential effects of taste or gut fill on oral testosterone consumption. Physiologically, it is difficult to achieve a rapid elevation in brain androgen with oral intake due to absorption across the gut and first-pass liver metabolism. Accordingly, to further evaluate the abuse potential of testosterone, we asked whether iv testosterone would support self-administration in male rats and hamsters [39]. Animals with chronic jugular cannulae were trained in an operant chamber to activate a nose poke to receive 50 ?g of testosterone intravenously. An inactive nose-poke hole served as a control. Within 5 days, rats expressed a significant preference for the active nose-poke hole (10.0±2.8 responses/4 h) over the inactive hole (4.7±1.2 responses/4 h). Similarly, during 16 days of testosterone self-administration intravenously, hamsters averaged 11.7±2.9 and 6.3±1.1 responses/4 h in the active and inactive nose-poke holes, respectively. By contrast, hamsters self-administering vehicle failed to develop a preference for the active nose-poke hole (6.5±0.5 and 6.4±0.3 responses/4 h). Compared with other drugs of abuse, testosterone reinforcement is modest. Nonetheless, these data support the hypothesis that testosterone is reinforcing.
3.6. Intracerebroventricular testosterone self-administration

Hamsters also self-administered testosterone (1 ?g in 1 ?l) directly into the lateral ventricles (Fig. 2), suggesting that testosterone reinforcement is mediated in the brain. Compared with iv androgen infusion, total androgen intake with intracerebrovascular self-administration was lower (27 ?g, icv vs. 342 ?g, iv), but response rates were higher (active hole: 39.8±6.0 nose pokes/4 h; inactive hole: 22.6±7.1 nose pokes/4 h). Replacing testosterone with vehicle extinguished responding in the active nose-poke hole, and reversing the active and inactive nose-poke holes increased responding in the previously inactive hole. However, reducing the testosterone dose from 1 to 0.2 ?g per 1 ?l of injection did not change nose poking.

4. Modifying androgen self-administration

To establish the potential for androgen reinforcement, our initial studies focused on gonad-intact adult male rats and hamsters because the majority of human AAS abusers are gonad-intact adult men. Individuals with low endogenous testosterone, including women, adolescents, and older men, are less likely to use AAS [40]. This suggests the possibility that endogenous androgens may enhance sensitivity to testosterone self-administration. However, many other factors, including undesirable peripheral masculinization and social pressures, affect AAS use in humans. Thus, we have begun to investigate how endogenous androgens, sex, and individual differences modify voluntary testosterone intake using icv self-administration in hamsters.

With oral testosterone self-administration at 3 mg/ml, average fluid intake in individual hamsters varied from 1.5 to 4.2 ml/3 h [38]. Likewise, with icv self-administration, mean responses on the active lever ranged from 64.6 to 20.2 nose pokes/4 h [39]. Many drugs of abuse show similar individual variation in self-administration. Saccharin preference has been used to predict ethanol self-administration [41]. In addition, rats selectively bred for brain stimulation reward or ethanol preference also consume more saccharin [42] and [43]. It has been suggested that a common neural mechanism mediates reinforcement for drugs, sex, food, and other natural and artificial reinforcers [28]. If so, one might anticipate similarities between saccharin preference and testosterone self-administration in individual males. We also compared testosterone intake and mating behavior, since mating requires gonadal steroids.

Initially, 21 male hamsters were prescreened for 0.1% saccharin intake and sexual behavior with a receptive female [44]. Thereafter, males were exposed to an aqueous solution of 3 mg/ml testosterone for 4 weeks using food-induced drinking. There were no significant correlations among saccharin, mating behavior, and testosterone intake. After chronic exposure to pharmacologic levels of testosterone, there was a significant increase in male sexual activity. However, the increase in ejaculations (153%) was modest considering the amount of androgen consumed (ca. 9 mg/day).

In addition to testing the relationship between androgen intake, saccharin preference, and mating, we investigated links between exercise and oral testosterone self-administration. Since AAS abuse in humans is strongly associated with athletes, and since hamsters voluntarily exercise in running wheels, we hypothesized that testosterone and exercise would show a reciprocal interaction in hamsters. That is, oral testosterone self-administration might enhance voluntary exercise, and exercise might stimulate androgen intake. This turned out to be only partly true: testosterone self-administration did not substantially enhance voluntary exercise, compared with vehicle administration to running males, and males with access to a running wheel self-administered no more androgen than did sedentary males [44]. Nonetheless, testosterone intake was positively correlated with wheel running in individual hamsters (Fig. 3). There is precedent for locomotor activity to forecast drug addiction. Rats selected for high locomotor response to a novel environment self-administered cocaine intravenously, while less active rats did not [45]. Likewise, high locomotor activity predicts iv self-administration of amphetamine [46].


The relationship between testosterone and exercise is complex. In athletes, particularly runners, prolonged endurance exercise is associated with decreased endogenous testosterone [47]. This appears to be mediated by the pituitary or hypothalamus, since secretion of luteinizing hormone is also reduced [48]. Conversely, strength training (weight lifting) causes a transient increase in circulating testosterone [49]. With exogenous androgen supplementation, muscle strength and size increase [50], and adipose tissue is reduced [51], and these effects are enhanced by exercise. Human steroid users report that they exercise more intensively during a steroid cycle and experience fewer injuries [8]. Thus, it appears that anabolic steroid use by athletes increases exercise, at least in part, through indirect effects on strength and reducing injury. Our observations in hamsters suggest that the inverse may also be true, namely that individuals motivated to exercise at high intensity may be predisposed to androgen reward.
5. Potential for androgen addiction
5.1. Comparison with other drugs

Based on the foregoing studies, it is evident that androgen reinforcement is not comparable to that of cocaine or heroin. Instead, it is likely that steroid reinforcement is similar to that of other mild reinforcers, such as caffeine, nicotine, or benzodiazepines. In this regard, rats in an operant chamber respond vigorously for iv heroin (28 ?g: 9 responses/30 min) [52] or cocaine (62.5 ?g: 130 responses/4 h) [53]. By contrast, whether by oral [38], iv, or icv self-administration [39], rats and hamsters show only a modest preference for testosterone (50 ?g, iv: 12 responses/4 h) [39]. Similar results have been observed for corticosterone self-administration by oral [36] and iv routes in rats [37] at comparable doses. Many other mild reinforcers do not support substantial self-administration. Caffeine (62.5?250 ?g) is not self-administered intravenously [53]. While rats do self-administer diazepam and nicotine intravenously, rates of operant responding are modest (500 ?g of diazepam: 25 responses/4 h [54]; 4 ?g of nicotine: 33 responses/4 h [55]). Moreover, rats prefer cocaine over nicotine in a two-lever choice test [55]. Although nicotine and benzodiazepines are mild reinforcers, it is remarkably difficult for many habitual users to quit. AAS may have similar effects.
5.2. Criteria for androgen addiction

Is testosterone addictive in humans? A number of studies suggest the potential for AAS addiction [11], [17], [18], [19], [21] and [22]. However, this issue is complicated by the interaction of androgens, athletic competition, and body image. Animal studies of testosterone self-administration and conditioned place preference demonstrate that testosterone is reinforcing in an experimental context where anabolic effects and athletic performance are irrelevant. Is testosterone addictive in animals? Many substances (e.g., sucrose) that are reinforcing are not necessarily addictive. Addiction is characterized by loss of control over use, such that subjects continue to seek out the drug despite adverse consequences [56]. Other criteria to establish addiction in animal studies include tolerance, withdrawal, and sensitization [57].

Recently, we noted several deaths during the course of icv testosterone self-administration [58]. Typically, hamsters died within several days of a ?binge? (up to 100 ?g) of testosterone intracerebroventricularly (Fig. 4). Symptoms of testosterone overdose included lethargy, low body temperature, and slow respiration. Self-administration of testosterone to the point of death suggests the potential for androgen addiction. However, this question requires additional study.

6. Neural mechanisms of androgen reward
6.1. Testosterone metabolites, steroid receptors, and the medial preoptic area (MPOA)

The brain is both the initial trigger for steroid production and also a principal target for steroid hormones. At the present time, the specific steroid signals, receptors, and brain sites of action for testosterone reinforcement are unknown. The reinforcing effects of testosterone may be androgenic, may be mediated by aromatization to estrogen, or may be sensitive to both androgens and estrogens. For many steroid-sensitive social behaviors in males, estrogen aromatized locally from testosterone is the active hormonal signal [59]. However, commonly abused AAS include both aromatizable and nonaromatizable androgens [12] and [13]. In the brain, testosterone can affect neural function through binding to classical androgen receptors, through classical estrogen receptors after aromatization to estrogen, or through nongenomic receptors [60]. Receptors for androgens are widely but selectively distributed in the basal telencephalon and diencephalon [61]. There is substantial overlap in the distribution of aromatase and receptors for androgens and estrogens [62], including both ? and ß forms of the estrogen receptor [63]. Finally, recent studies have demonstrated rapid effects of androgens and estrogens in brain regions that possess few classical receptors [64]. These steroid actions are thought to be mediated by nongenomic receptors. There is the additional possibility that testosterone reinforcement may be mediated by a combination of classical and nongenomic receptors.

Testosterone reinforcement does not necessarily follow the same mechanisms previously established for steroid effects on social behavior. In terms of brain site(s) for testosterone reinforcement, the MPOA is a key site for organization of male sexual behavior [59]. MPOA has abundant steroid receptors, and testosterone implants in MPOA restore sexual activity in long-term castrates. The time course of these steroid effects is slow: mounting behavior persists for weeks after orchidectomy, and extended steroid exposure is necessary to restore mating in long-term castrates. However, injections of testosterone into MPOA of male rats fail to induce conditioned place preference [65]. This suggests that other brain regions are important for androgen reinforcement.
6.2. Dopamine and Acb

The mesolimbic dopamine system is an essential brain circuit for motivation and reward (see Ref. [57] for review). Dopamine is released into Acb from neurons in the ventral tegmental area (VTA) during natural rewards, such as food or sex [66]. Drugs of abuse act on the mesolimbic dopamine system to increase dopamine release (amphetamine, opiates), or reduce dopamine reuptake by nerve terminals in Acb (cocaine) [67]. Dopamine release from VTA neurons and dopamine receptors in Acb are critical for these effects. Selective lesions with 6-hydroxy-dopamine in VTA or Acb attenuate the reinforcing properties of food, sex, and drugs of abuse [68]. Moreover, rats will self-administer many drugs directly into Acb or VTA [69] and [70].

Testosterone injected directly into Acb induces conditioned place preference [32]. It is probable that this effect is mediated by nongenomic steroid receptors because Acb and VTA have few steroid receptors [71]. As with other drugs of abuse, dopamine is likely to be a key neurotransmitter for testosterone reinforcement: conditioned place preference induced by systemic testosterone injection is blocked by D1 and D2 dopamine receptor antagonists [72]. However, we did not observe signs of stereotypy or locomotor sensitization following systemic injection of 800 ?g/kg testosterone [73]. Likewise, testosterone injections did not alter stereotypy induced by the dopamine agonist, apomorphine (1.2 mg/kg). Together, these data suggest that although testosterone reinforcement may ultimately enhance dopamine activity in Acb, the mechanisms may be distinct from those of cocaine or other stimulants.

There is a precedent for steroid effects on the mesolimbic dopamine system: progesterone metabolites are rewarding when delivered directly to VTA [74]. This effect does not require progesterone receptors. Instead, progesterone metabolites bind to GABA/benzodiazepine receptors and open Cl? channels [75]. Androgens also modulate activity of the GABA/benzodiazepine receptor [76], and previous studies have suggested parallels between testosterone and sedative?hypnotics. Like benzodiazepines, testosterone and its derivatives have anxiolytic and analgesic effects, as demonstrated with open field, tail flick, paw lick, defensive burying, and social interaction tests in rats [77] and with an elevated plus maze in mice [78]. These effects can be rapid: within 30 min of injection, testosterone (800 ?g) increased time spent at the distal ends of the open arms of an elevated plus maze (57.1±14.9 s/5 min vs. 33.7±9.7 s/5 min at baseline) [73].
7. Behavioral consequences of androgen abuse

From a public health perspective, there is concern that AAS may have a negative impact not only on steroid users, but also on those around them. Through anecdotal reports of violent behavior in AAS users, ?roid rage? is widely accepted in the popular media: a search on Google yielded 5200 results. However, few studies in humans have determined if androgens at pharmacologic doses cause uncontrolled aggression. From a clinical trial of nonathletes receiving 600 mg of testosterone enanthate per week, Tricker et al. [79] concluded that testosterone does not increase angry behavior. Nonetheless, AAS may provoke violence in vulnerable individuals (competitive athletes or those with preexisting psychopathology). In rats and hamsters, AAS can modify both sexual and aggressive behaviors. In particular, testosterone propionate significantly increased sexual behavior and aggression in peripubertal and adult male rats and hamsters [80], [81], [82], [83], [84], [85], [86] and [87]. AAS appear to stimulate aggression in male hamsters by increasing the expression of arginine vasopressin in the anterior hypothalamus [86] and [87]. In female rats, AAS advance vaginal opening but interrupt estrous cyclicity [88], [89], [90], [91] and [92]. Steroid effects on behavior are transient because symptoms resolve within several weeks after steroid withdrawal [90] and [93].
8. Fatal consequences of androgen abuse

Many AAS users are aware of the negative cosmetic side effects of high-dose steroid use, including acne and balding, gynecomastia (?***** tits?), and testicular atrophy in men, and clitoromegaly, facial hair, and lowering of the voice in women [12], [13], [16], [94] and [95]. In addition to causing infertility in both sexes, AAS use at pharmacologic doses can produce serious health consequences, including myocardial infarction, cardiomyopathy, stroke, behavioral disturbances, and hepatic tumors and lesions [12], [16], [96], [97], [98], [99] and [100]. Although Lyle Alzado, former NFL defensive lineman, attributed his fatal brain tumor to long-standing steroid abuse, links with cancer have not been established. However, several small studies and case reports describe fatalities with AAS use due to hepatic, cardiovascular, and psychiatric/behavioral dysfunction.
8.1. Deaths from AAS abuse: peripheral effects

In the treatment of anemia, oral administration of the C17 alkylated androgen oxymetholone has been used to stimulate erythropoiesis. However, chronic androgen exposure can produce liver dysfunction, including peliosis hepatis, hepatocellular carcinoma, and cholestasis [16]. Nonetheless, symptoms often resolve when steroid use is discontinued, and fatalities from steroid-induced liver dysfunction are infrequent in otherwise healthy individuals. Among athletes and body builders, AAS abuse can occasionally precipitate arrhythmia, stroke, and myocardial infarction with sudden death in young adults [16]. Many of these deaths occur during exercise, with cardiac hypertrophy noted at autopsy. The anabolic effects of androgens include hypertrophy of cardiac myocytes. Without compensatory capillary growth, the myocardium in AAS users may become relatively ischemic [101]. Hypertrophic cardiomyopathy is the most common cause of cardiac sudden death during exercise in young (<35 years) athletes [102]. Androgen-stimulated thrombosis could also contribute to heart failure in AAS users [103].
8.2. Deaths from AAS abuse: central effects

In clinical reports, there is precedent for deaths due to androgen effects on brain and behavior: homicide, suicide, accident, and polydrug abuse [104]. Heroin addicts face similar risks [105]. AAS contribute to homicide and suicide by precipitating behavioral changes such as increased risk taking, lack of impulse control, aggression, and depression [104]. In addition to indirect behavioral effects of androgen abuse, AAS use may be fatal in combination with other drugs. In particular, steroids may lower the threshold for opiate overdose [104]. Together, these findings suggest that AAS, alone or in combination with other drugs, can be fatal