Anabolic androgenic steroids are often used by athletes and body builders in an effort to improve their performance or appearance, in many cases starting in high school^(9,10,16,26). Athletes and body builders often take as many as five different steroids simultaneously, usually including testosterone, injectable esters of testosterone, and analogues that can be taken orally ^(17,23). The combined doses of these hormones typically total 10-40 times normal androgenic maintenance levels ^(15,17), and often the steroids are taken in combination with growth hormone and IGF-1 ⁽¹²⁾. Usually the steroids are taken in cycles, starting several weeks or months prior to competition, followed by a washout period of weeks or months. This pattern of use can continue for years ⁽¹⁵⁾.

The potential pathological consequences of exposure to such high levels of steroids have been of continuing concern for some time now^{(11,12,14,17)}. In an overview of the state of our knowledge about this subject, Friedl ⁽¹¹⁾ concludes that “.... an athlete would be foolish to conclude that there is a safe way to use anabolic steroids; although no disease of androgen excess has ever been described for men, the long-term consequences of androgen supplementation have not been investigated and are simply unknown.” Reviews of case histories suggest that men using steroids may show a marked increase in liver disorders, including liver tumors, a decrease in fertility, and a shift in the serum lipoprotein profile in a way that could indicate an increased risk of heart disease^(11,12,19).

The only animal studies directed at the potential pathological consequences of exposure to pharmacological amounts of anabolic-androgenic steroids have involved short term experiments of 6 weeks or less^(2,3,27). The object of the present study was to assess the effect of exposure to anabolic steroids on the life span and cause of death of male mice using the numbers, kinds, and relative amounts of steroids used by athletes and body builders. Specifically, adult male mice were exposed for 6 months to a combination of four steroids, including testosterone, an injectable ester of testosterone, and two 17-alkylated steroids tailored for oral use. In line with the doses taken by humans, the four hormones were given to mice at doses that totaled either 5 or 20 times normal androgenic maintenance levels for male mice. The animals were observed for 1 yr following the termination of exposure to steroids.

MATERIALS AND METHODS

Animals and experimental design. CF-1 male mice were purchased from Charles River (Portage, Michigan) at 6 wk of age. They were housed one per cage from the time of arrival until their death. At 8 wk of age, after adult reproductive capacity had been achieved, the animals were delegated to three experimental groups that were matched for the average and range in body weight. The males in two of these groups were exposed for 6 months to a combination of four anabolic-androgenic steroids at doses that were either 5 or 20 times greater than those required to maintain normal-sized seminal vesicles in castrated CF-1 male mice. Using an inhalent anesthesia, metofane(methoxyflurane, Pitman-Moore, Mundelein, IL), the capsules were placed under the skin on the back of the mouse, closing the small suture with a single wound clip that was later removed. The third group of males was given empty capsules. There were 43 to 50 males in each of the three treatment groups.

Our experimental protocol called for all animals to be killed with CO₂ when judged to be on the verge of death using criteria to be described. All organs were to be saved, and after staining with hemotoxylyn and eosin, they were to be subjected to histological examination. Thus age at death and the pathological conditions present at that time were to be recorded for all animals until 1 yr after exposure to steroids was terminated, when the survivors were 20 months old. These objectives were met in 46 of the 59 males dying before the experiment was terminated. As will be detailed later, 13 males died relatively suddenly, without any noticeable indications of impending mortality beforehand. Autolysis prevented detailed histological examination of these 13 animals. Thus, the age at which they died was included in our analysis of the effect of steroids on life span but not in our study of the pathological conditions present at the time of death.

The animals in this study were housed singly in 29 × 18 × 12 cm polyethylene cages on pine shavings and maintained on a 14:10 light:dark cycle. They were fed Purina Formulab 5008 (St. Louis, MO) and given waterad libitum.

This experiment was done in accordance with the policy statement of the American College of Sports Medicine on research with experimental animals⁽¹⁾ and in accordance with protocol number 06952201 of the IACUC of the University of Texas at Austin.

Steroid doses. The rationale and procedures used to establish the hormones, doses, and method of delivery used in this study have been presented in detail elsewhere ⁽⁵⁾. As noted earlier, however, one object of this experiment was to expose male mice to a combination of steroids that spanned the classes of steroids used by athletes and body builders. We combined testoserone, testosterone cypionate, which is an injectable ester, methyltestosterone, which is an orally-taken 17-alpha-methyl derivative, and norethandrolone, which is an orally-taken 17-alpha-ethyl derivative with a modified ring structure ^(23,24). A second object was to give this combination of hormones in relative amounts typical of those used by athletes and body builders. We chose two doses, 5 and 20 times normal androgenic maintenance levels in mice, which are typical of the doses used by athletes ^(15,17). The doses of steroids given in this experiment were based on their androgenicity as determined by bioassay, the weight of the seminal vesicles of a castrated male CF-1 mouse. It should be noted that the androgenic effects of steroids may not be the same as their anabolic effects at pharmacological levels ⁽¹³⁾.

To ensure continuous delivery of these hormones over a 6-month period, each hormone was packed in its own Silastic capsule. Silastic capsules are characterized in terms of three dimensions: inner diameter, outer diameter, and length. The rate of release of a hormone from a Silastic capsule is proportional to the length of the capsule for any given set of diameters, and the rate of release remains constant over periods of time measured in months or years depending on the hormone and the size of the capsule^(4,8). Typically, dose response curves using Silastic capsules are produced by varying capsule length while keeping the inner and outer diameter of the capsules constant.

The sizes of capsules used in this experiment were chosen with two goals in mind: to combine the four hormones noted earlier in such a way that each would contribute one-fourth of the total androgenicity at a given dose, and second, to keep capsule length as short as possible to minimize discomfort. To accomplish these goals it was necessary to vary all three dimensions of the capsules depending on the dose desired and the hormone to be administered.

Choice of capsule sizes proceeded in two steps. First, the maintenance“dose” of each hormone was determined for capsules of different diameters by systematically varying capsule length. Maintenance dose was defined for each hormone as the capsule size required to maintain the seminal vesicles of a castrate male mouse at a weight typical of intact males. In separate studies for each of the four hormones, adult CF-1 males were castrated, immediately given hormone implants of various diameters and lengths, and then killed 2 wk later, at which time the stripped weight of the seminal vesicles was obtained. For comparison, each study also included a group of castrated males given only blank capsules to establish a baseline and a group of intact males to establish the normal weight of the seminal vesicles(see below).

The second step was to determine if one could expose mice to all four hormones simultaneously with predictable results based on the capsule size of each. Castrated males were given the following four implants: 2.5 mm of a 0.062-inch (inner diameter) × 0.125-inch (outer diameter) capsule containing testosterone; 5 mm of a 0.062-inch × 0.125-inch capsule of testosterone cypionate; 2.5 mm of a 0.025-inch × 0.047-inch capsule of methyltestosterone and 7.5 mm of a 0.062-inch × 0.125-inch capsule of norethandrolone. In each case the capsule length was one-fourth that shown to produce a maintenance dose when the hormone was administered alone in a capsule of the stated diameter. At autopsy 2 wk later, the mean seminal vesicle weight of the animals given the four hormones as a combination was 88.2 ± 4.0 mg compared with 84.8 ± 5.7 for intact males and 23.0± 2.3 for castrates given blank implants (N = 8 in all cases). This combination of capsule sizes was defined as the maintenance dose for the combination of four hormones.

In the experiment reported here, the 5-fold maintenance dose combined the following: 12.5 mm of a 0.062-inch × 0.125-inch capsule containing testosterone; 5 mm of a 0.078-inch × 0.125-inch capsule of testosterone cypionate; 12.5 mm of a 0.025-inch × 0.047-inch capsule of methyltestosterone; and 7.5 mm of a 0.078-inch × 0.125-inch capsule of norethandrolone. The small capsule containing methyltestosterone had to be replaced after 4 months. The 20-fold high dose combined: 10 mm of a 0.078-inch× 125-inch capsule of testosterone; 20 mm of a 0.078-inch × 0.125-inch capsule of testosterone cypionate; 10 mm of a 0.062-inch × 0.125-inch capsule of methyltestosterone; and 30 mm of a 0.078-inch × 0.125-inch capsule of norethandrolone.

Indications of impending death. Two kinds of data were of interest in this study: age at death and cause of death. Ideally, constructing an accurate mortality curve requires that animals actually be found dead in their cages. To avoid as much suffering as possible and to collect tissues suitable for histological examination, however, one needs to kill the animals rather than allowing them to die. To accommodate these diverse needs as closely as possible, we used a two-stage process to determine when to kill an animal. First, early in the study each animal was examined closely once a week. As the animals aged, this schedule was changed to a twice weekly examination and, later, to a thrice weekly examination. During these examinations, animals were judged to require more frequent observation if they showed one or more of the following symptoms: loss of body weight, pale eyes, a hunched posture when at rest, a distended abdomen owing to kidney or liver tumors or, conversely, sunken sides. Animals showing these signs were observed once or twice a day thereafter and killed when they appeared moribund as defined by the following symptoms: a poor or delayed reaction to physical stimulation, poorly coordinated (wobbly) locomotion, and/or labored breathing. Determination of the time to kill an animal was done blind with regard to treatment usually by mutual agreement of two people, but on occasion the decision was made by only one or the other person alone.

Statistical analysis. To assess the effect of steroid exposure on life span, we compared the number of animals in the three experimental groups that died before the experiment was terminated. This occurred when the survivors were 20 months old, 1 yr after exposure to steroids ceased. The comparison was tested statistically by X²⁽²⁰⁾.

RESULTS

As shown in Figure 1, by 20 months of age 52% (26/50) of the males treated with the high dose of steroids had died or been killed compared with 33% (15/46) and 12% (5/43) of the males given the low dose and the blank capsules, respectively (P < 0.001; X²). Of the 59 animals that died, only two did so before the animals' steroid capsules were removed. Both had been given the high dose of steroids, and both died of kidney failure caused by a glomerulonephropathy.

As noted earlier, 13 males died before they could be killed by the experimenter: three control males, six low dose males, and four males given the high dose of steroids. Table 1 presents a survey of the pathological conditions seen at autopsy or revealed later by histological analysis of the other 46 males. Most of the steroid-treated animals dying before 1 yr of age suffered from the glomerulonephropathy mentioned above. Most of the steroid-treated mice dying after 1 yr of age displayed tumors, most often in the liver or kidney, or both. As noted in Table 1, hepatocytic carcinomas were relatively common in the liver of steroid-treated males, sometimes with secondary invasion of other organs, often the lungs. Peliosis hepatis, often with internal hemorrhaging, was also common in the steroid-treated males. Mostly in the kidney we saw mesenchymal tumors of the fibrosarcomatous, leiomyosarcomatous, or rhabdomyosarcomatous form.

Again, as shown in Table 1, four steroid-treated individuals, three given the low dose and one the high dose, showed pronounced left ventricular myocardial hypertrophy, usually associated with fibrosis and stenosis of the left atrio-ventricular valve. One male given the high dose of steroids and one given the low dose showed multifocal myocardial fibrosis. Three males given the high dose of steroids showed pulmonary adenomas, and another high dose male showed a bronchiolar adenocarcinoma. One control male, one male given the low dose of steroids, and five males given the high dose of steroids suffered from lymphosarcomas, typically involving the thymus, spleen, and lymph nodes.

DISCUSSION

The object of this study was to use the male laboratory mouse as an animal model with which to suggest potential pathological consequences of steroid abuse by male athletes and body builders. The degree to which the results are relevant to humans depends on two factors: the degree to which the parameters of exposure in this experiment mimic those occurring in humans and the degree to which the specific pathological effects seen in mice mimic those seen in humans.

As detailed earlier, the numbers, kinds, and relative doses of steroids to which our mice were exposed are quite like the numbers, kinds, and relative doses of steroids to which athletes and body builders expose themselves. Indeed, the low dose we gave our mice is quite a bit lower than the doses typically used by athletes and body builders. As noted earlier, athletes and body builders usually take steroids in doses totally 10-40 times maintenance level ^(15,17), while our mice were given doses of either 5 or 20 times maintenance level (for mice).

On the other hand, the duration of exposure to which our mice were subjected may not be typical of that occurring in humans, depending on how one evaluates duration in relation to life span. Most male mice live to 30 months(and a few are still functioning sexually at 3 yr of age)⁽⁶⁾. Thus 6 months of exposure to steroids is about one-fifth of a male mouse's life expectancy. This is undoubtedly longer than the duration to which most athletes and body builders expose themselves. Constant exposure over this period of time, without periodic washout periods, is also atypical of human use. Given all of these comparisons, we are probably only safe in concluding that the exposure parameters used in this experiment may be generalized for extreme steroid abusers.

Some of the pathological conditions seen in steroid-treated mice are similar to those seen in athletes and body builders and some are not. In the latter category are the two kinds of kidney damage seen in mice-the direct toxic effect on the glomeruli that killed many mice during the first year of their lives and the development of kidney tumors at later ages. There is only one report of a kidney tumor in a human athlete ⁽¹⁸⁾ and no suggestion of a direct nephrotoxic effect in either athletes^(11,12) or nonathletes given higher-than-maintenance levels of androgens in clinical trials (e.g., in tests of the utility of androgens to act as male contraceptives⁽²¹⁾).

In contrast, the kinds of liver damage seen here in mice, most typically hepatocytic carcinoma and peliosis hepatis, are some of the most common pathological effects seen in athletes and body builders⁽¹⁹⁾. The fact that suprapharmacological amounts of anabolic steroids can damage the liver of rodents in a variety of ways, including the induction of tumors, has been reported previously⁽²²⁾. The results of the present experiment demonstrate that the kinds of liver damage seen in athletes can be induced in mice using the same numbers, kinds, and relative amounts of steroids used routinely by athletes and weight lifters. Thus, our results with mice support the hypothesis that steroid abuse can result in serious liver damage. Friedl's⁽¹¹⁾ analysis of case histories suggests that 17-alkyl-substituted androgens cause liver damage while androgen esters do not. Thus, ultimately the effects seen in mice in this experiment may be traceable only to exposure to methyltestosterone and norethandrolone.

Perhaps the most important result of the present study is the demonstration that exposure to steroids produces a broad array of pathological effects that do not appear until long after exposure to steroids ceases. There is little comparable data for humans ⁽⁷⁾. Widespread use of steroids did not occur until the 1970s and, probably more germane, the practice of“stacking” or combining several analogues of testosterone at suprapharmacological levels did not become common until the 1980s⁽²⁵⁾. Thus, the delayed effects of steroid abuse seen here in mice and the consequent dramatic effect on life span may ultimately prove to be a concern for athletes and body builders abusing steroids regardless of specific pathological condition.