Steroid-Induced Mental Disturbances

Original Author: Paul Perry, Ph.D, BCPP
Latest Revisers: Paul Perry, Ph.D, BCPP, Brian C. Lund, Pharm.D.
Creation Date: 1996
Last Revision Date: April 2004
Peer Review Status: Internally Peer Reviewed

Corticosteroids

Introduction

Cases of mental disturbances associated with corticosteroid use have been documented since the early 1950's. The frequency of occurrence varies from study to study. In 1972, the Boston Collaborative Drug Surveillance Program (BCDSP) (1972) reported 21 cases (3%) among 718 patients receiving prednisone. This is a conservative figure since the study only included patients who were emotionally stable and free from mental disorders before treatment. In contrast, there is a report of a 57% incidence (16/28) of mental disturbances after treating a group of systemic lupus erythematosus (SLE) patients with prednisone (Sergent et al 1975). This high incidence is probably exaggerated by the disease itself as it is well known that patients with SLE may exhibit psychoses even without drug therapy. Additionally, the massive doses of prednisone, 100-500 mg/d, were a significant contributing factor. Lewis and Smith (1978) reviewed 13 studies containing 2,555 patients who were treated with corticosteroids. In these studies, the incidence of severe psychiatric syndromes ranged from 1.6% to 50% with a weighted mean of 5.7%.

ACTH, cortisone and prednisone are the three corticosteroids most commonly associated with mental disturbances. Their apparent preponderance may be due to the fact that they have been the most commonly used corticosteroids. Others, such as triamcinolone, dexamethasone, methylprednisolone, prednisone-21-stearoylglycolate, (Rosenberg et al 1976) and even beclomethasone (Kreus et al 1975) inhalation have been associated with mental disturbances. In general, all corticosteroids have at least some potential to induce mental disturbances.

Biological Effects

Wolkowitz (1994) reviewed a series of studies that illucidated the biological effects of exogenous corticosteroid administration. Prednisone 80 mg/d for 5 days was administered to healthy human volunteers. This resulted in significant decreases in plasma levels of ACTH, cortisol, and MHPG (NE metabolite). In the CSF, significant decreases of ACTH, beta-endorphin, beta -lipotropin, somatostatin-like immunoreactivity, and NE occurred. Significant slowing of brain wave activity, i.e., increased theta wave activity, was observed on EEG. Memory testing suggested increased errors of commission.

Cause and Effect Relationship

Due to the lack of control cases for comparison, anecdotal case reports alone are insufficient to establish a definite causal relationship between corticosteroid treatment and mental disturbances.

Positive Cause-and-Effect Relationship
Marx and Barker (1967) compared two groups of patients with ulcerative colitis or ileocolitis. Among 50 patients treated with corticosteroids, six (12%) developed acute psychoses while only two (2.6%) of the 80 control patients did. Because of this significant difference it was concluded that there was an increased likelihood of mental disturbances when patients were treated with corticosteroids. However, it is difficult to validate this conclusion since the two groups were not matched for age, sex, clinical condition, and past psychiatric history.

Cade et al (1973) found that corticosteroids precipitated mental disturbances in patients with lupus nephritis. Fifty-five patients were randomly assigned to four treatment groups. Group 1 consisted of 15 patients treated with prednisone alone, Group 2 of 13 patients treated with azathioprine alone, Group 3 of 13 patients treated with a combination of prednisone and azathioprine, and Group 4 of 14 patients treated with heparin and azathioprine. The dosage of prednisone used in Groups 1 and 3 was 60 to 100 mg/d for 6 months, and then it was slowly reduced to the lowest possible dose. Of patients not receiving prednisone (Groups 2 and 4), none developed psychoses, whereas 9 out of 28 patients receiving prednisone developed overt psychoses. This difference was statistically significant (p<0.001). The authors concluded that although prednisone may not be the only factor in producing mental disturbances, it is likely that it plays an important role.

Negative Cause-and-Effect Relationship
Smyllie and Connolly (1968) retrospectively evaluated 550 patients treated with corticosteroids for respiratory diseases. As controls, he used 499 patients who had not received corticosteroids for at least one year previously and never continuously for more than one month. The two groups were matched approximately for disease, age, sex, history of psychiatric illness, and year of entry into the hospital. Only serious psychoses that required psychiatric consultation and/or treatment were considered as mental disturbances. Analysis showed that only 1.8% (10 patients) of the treated group developed serious mental disturbances while a 3.2% incidence (16 patients) was seen in the control group. The author concluded that there was no extra risk of mental disturbances incurred by corticosteroid therapy. However, it must be kept in mind that 94% of the corticosteroid-treated patients were taking the equivalent of less than 20 mg per day of prednisone. As can be seen in the BCDSP prednisone data below this is an extremely relevant observation.

Dosage and Duration of Corticosteroid Treatment

The Boston Collaborative Drug Surveillance Program (1972) (BCDSP), however, convincingly demonstrated a statistically significant correlation between dosage and incidence of corticosteroid-induced mental disturbances. Among the 718 patients treated with prednisone, psychiatric symptoms were recorded in 1.3% (6/463) of the patients receiving 40 mg or less per day; in 4.6% (8/175) of those receiving 41-80 mg/d; and in 18.4% (7/38) of those receiving greater than 80 mg/d. A test for trend was statistically significant. They also found that the mean dosage of prednisone in the 21 patients showing mental disturbances was 60 mg/d, which was significantly higher than the 31 mg/d in the patients without adverse reactions. These findings not only showed that corticosteroids could indeed cause mental disturbances, but also that the incidence was dose-related. Of the 21 patients, 13 (62%) were described as psychotic i.e. hallucinations, delusions and/or violent, 6 (29%) as manic and 2 (10%) as depressed. In the BCDSP (1972) study, only patients taking prednisone were monitored and all patients were free from psychiatric illness before treatment. By controlling these two variables the study subjects were more homogeneous. Thus, increasing the dose of corticosteroids increases the risk of mental disturbances.

Time of Onset

The time to onset of mental disturbances varies widely from patient to patient. The quickest onset was within the second day of corticosteroid treatment (Clark et al 1953). Although one patient was taking the drug for 95 days prior to symptoms (Clark et al 1952), most required a few days to one or two weeks before the symptoms appeared. Of 37 cases where time of onset of the psychosis could be determined for the corticosteroid, 62% of the psychiatric disturbances < 8 days. There are also cases where the patients were unaffected during previous courses but developed mental disturbances during the subsequent courses (Clark et al 1953). Thus, it is impossible to generalize about the time of onset. Moreover, dosage does not appear to affect the time of onset.

Reactions

Affective symptoms are the most frequently observed reactions to corticosteroids. The euphoric effects vary in degree and appropriateness. Some patients may only show cheerfulness and well-being which is appropriate to the improvement of their illnesses. Others may become inappropriately elated, even to exhibit hypomania. Depression is often observed in patients treated with corticosteroids. The patient may simply be pessimistic about the illness or the physical side effects of the drug(s), or may develop feelings of guilt and even become suicidal.

Ling et al (1981) surveyed 55 case reports of steroid-induced psychiatric disturbances. Twenty-two cases (40%) had symptoms of depression only, 17 (31%) presented with only manic symptomatology while 6 (11%) experienced alternating depressive and manic symptoms. In 9 cases (16%) delusions, hallucinations, stupor and/or catatonia occurred alone. The remaining case (2%) was characterized as a state of agitation and anxiety. In addition, it was quite common for the delusions and/or hallucinations to occur simultaneously with the affective symptoms. Of the 55 case reports, 23 (42%) presented as either depression or hypomania accompanied by psychotic symptoms. Thus, because of the mixture of psychiatric symptoms, one cannot routinely describe a steroid-induced psychosis as a well-defined delirium, depression, or manic episode. It should be noted that of the 55 case reports surveyed by Ling et al, 32 (58%) involved ACTH, either alone or in combination with cortisone or in one case prednisone. Excluding ACTH from the available case reports in the analysis of the presentation of corticosteroid-induced psychiatric disturbances produces a slightly different cross-sectional presentation. Perry et al (1984) reviewed a total of 43 cases of corticosteroid-induced psychosis. Twenty-two (51%) presented primarily as mania or hypomania, 12 (28%) as depressed or dysphoric states, while two cases (5%) presented as a combination of dysphoria and hypomania. Thus manic rather than depressive symptoms probably are more common with the corticosteroids, although mixed affective symptoms still dominated the psychiatric presentation in the corticosteroid-disturbed patient. Of the 37 cases demonstrating an affective component 26 (70%) were accompanied by hallucinations and/or delusions. Of the six cases not involving an affective component, four presented primarily as hallucinations and/or delusions while two cases of delirium were described.

The affective mental status changes observed in patients with a corticosteroid-induced Cushinoid state are consistent with the psychiatric changes observed with Cushing's syndrome. Haskett (1985) obtained a longitudinal psychiatric history from 30 Cushinoid patients. Twenty-five (83%) met Research Diagnostic Criteria for an episode of affective disorder during the course of their endocrine disturbance. The psychiatric diagnoses of the 30 patients could be divided into three major groups: 16 patients with depressive disorder only (unipolar affective disorder), nine patients with manic, hypomanic, or mixed affective disorders (bipolar affective disorder), and five patients with no psychiatric diagnoses. Schizophrenic symptoms were not obvious in any patients, although two patients met criteria for psychotic depression. Affective syndromes occurred in patients with adrenocortical tumors (6/6) as well as in those with ACTH-dependent Cushing's syndrome.

A number of uncontrolled trials and one controlled trial have examined the potential effectiveness of antiglucorticoid drugs in the treatment of major depression, including some hypercortisolemic depressed patients. The drugs included the cortisol synthesis inhibitors ketoconazole (Nizoral), aminoglutethimide (Cytadren), and metyrapone (Metopirone) and the corticosteroid receptor antagonist mifepristone (RU-486) (Wolkowitz and Reuss 1999). Ketoconazole has been studied to the greatest extent. The largest controlled randomized placebo controlled trial using 400-800 mg/d for 4 weeks included 20 patients diagnosed with major depression (DSM-IV). Only 8 of the 20 patients were hypercortisolemic. It was only in these patients that ketoconzaole was more effective than placebo in reducing the symptoms of depression according to the HAMD change scores separated from placebo in effectiveness in the (Wolkowitz et al 1999).

Age

A breakdown by age of the 55 anecdotal cases reviewed by Ling et al (1981) showed that the majority of patients suffering from corticosteroid-induced mental disturbances were between the ages of 21 and 60. It is misleading, however, to look at age as a single variable because the prevalence of diseases requiring corticosteroid treatment can also be greater in certain age groups. For example, rheumatoid arthritis, systemic lupus erythematosus, and pemphigus occur more often in middle-aged people or sometimes in young adults. There are few people over 60 suffering from corticosteroid-induced mental disturbances because patients with diseases that are serious enough to require corticosteroid treatment usually have a shorter life expectancy. Other diseases such as asthma may occur in children, but they are usually not in such stage as to require systemic treatment with corticosteroids. Treatment may be required as the child grows older and the disease progresses.

Although the age range can vary from 4 months to 84 years, it is rare, that very young or very old people require corticosteroid treatment. Accordingly, there is no reason to suspect that certain age groups are at a higher risk of developing corticosteroid-induced mental disturbances than others.

Sex

Hayreh and Watson (1970) noted that two females in a group of 13 females and 14 males treated with prednisone-21-stearoylglycolate (Sintisone) developed acute psychoses with suicidal ideation.

Bunim et al (1955) treated 64 patients suffering from rheumatoid arthritis with corticosteroids; 59 of them (32 females and 27 males) had adverse effects. Depression occurred in four of the women, but in none of the men.

Nielsen et al (1963) reported 12 cases of depression or pronounced restlessness after corticosteroids were given to 50 patients (15 men and 35 women) to treat rheumatoid arthritis. All 12 cases were female.

Cade et al (1973) treated four groups of patients for lupus nephritis, two with prednisone and two without. The corticosteroid groups consisted of 28 patients, 22 females and 6 males. Eight females showed mental disturbances after the therapy while only one male patient was mentally disturbed.

Summing of the above numbers shows the occurrence of steroid psychosis to be 15-fold more common [26/108 (24%) versus 1/62 (1.6%] in females than in males. Thus, similar to affective illness in general, it is more likely to see corticosteroid mental status changes in women than in men.

Past History of Mental Illness

Lewis and Smith (1983) described 14 cases of steroid-induced psychiatric syndromes. None of the 14 cases had a past history of psychiatric illness unrelated to corticosteroid therapy. Six (43%) were thought to have evidence of a premorbid personality disorder. They noted that among 41 cases of steroid intoxication they were able to identify in the literature, 17% had a prior history of psychiatric illness unrelated to corticosteroids. Fifty-two percent of 29 cases were reported to have had an abnormal premorbid personality. Thus, based on these minimal data the authors concluded that it was not possible to determine whether past psychiatric illness or premorbid personality disturbances are risk factors for the development of corticosteroid mental status changes.

Treatment of Corticosteroid-Induced Mental Disturbances

Most of the milder forms of corticosteroid-induced mental disturbances disappeared regardless of dosage reduction or tapering of the drugs. Nevertheless, there are more severe cases that require specific treatment since corticosteroids cannot be abruptly discontinued. Before the availability use of the antipsychotics, the treatment was sedation, or ECT, as well as hospitalization, for those with suicidal ideation. At present, antipsychotics are effective and are the drugs of choice in acutely psychotic patients. ECT is indicated in depressives who do not respond to discontinuation of the drug or dose reduction (Sutor et al 1996). Because of their anticholinergic activity, it is preferable to avoid tricyclic antidepressants. Of the SSRIs, paroxetine has some anticholinergic activity and the other agents of this class are preferred. This is supported by a series of steroid psychosis cases, in which 4 of the patients were treated with tricyclics (Hall et al 1979). All four patients experienced deterioration in their mental status on tricyclics, but improved once the tricyclic was replaced with a antipsychotics. Since potent anticholinergic tricyclics were utilized to treat these patients, it was hypothesized that an anticholinergic delirium was layered on top of the corticosteroid-induced depression. Thus antidepressants without anticholinergic activity such as fluoxetine and trazodone are recommended. For manic symptomatology, lithium is the drug of choice.

Patients with chronic illnesses that require periodic high-dose corticosteroid treatment and who have experienced a corticosteroid mental status disturbance in the past should be treated prophylactically with lithium. Falk et al (1979) found lithium prophylaxis useful in preventing corticotrophin-induced psychoses. During a 31-day course of corticotrophin therapy for acute exacerbations of multiple sclerosis or retrobulbar neuritis, 27 patients were empirically treated with lithium carbonate doses sufficient to maintain the serum levels between 0.8-1.2 mEq/L. Mental disturbances occurred in none of these patients. However, in a comparable group of 44 patients treated with corticotrophin but without lithium, six (14%) became psychotic. Thus, the short- term lithium prophylaxis may be considered an alternative in patients with a prior history of corticosteroid-induced mental disturbance who require additional courses of corticosteroid therapy.

Dependence and Withdrawal Effects

It is important to note that corticosteroid withdrawal can also induce mental disturbances. As a result of the euphoric effect, it is possible for some patients to be psychologically dependent on the corticosteroids (Kimball 1971, McCawley 1965, Morgan et al 1973). Withdrawal reactions typically involve depression, but agitation, anxiety, and psychotic reactions have been observed. Because both use and withdrawal of corticosteroids may lead to mental disturbances, time of onset and course of the mental disturbance should be examined closely during assessment.

Anabolic Steroids

Epidemiology

The prevalence of AS use has been reported in several populations. The most recent estimates report a range 4-12% of U.S high school boys have used AS at some time in their life (Yesalis and Bahrke 1995). Sullivan et al (1998) reported that 65-84% of adolescent AS users were participants in organized athletics. On the other hand, a surprising report showed that 3.2% of Modesto, California 7th grade girls have admitted to using AS (Yesalis and Bahrke 1995). These numbers are worrisome considering the adverse growth suppressing effects of AS in young adults (Buckley et al 1993). Yesalis et al (1990) reported 29.3% of college football players and 20.6% of track-and-field male athletes reported AS usage. Among college football players, lifetime prevalence of AS use was 29.3%. Some of the highest estimates have come from Yesalis and Bahrke (1995); reporting that 78% of track-and-field athletes in 1972 had prior steroid use. In a recent survey of former Olympians, they found that 75% of medallists and 63% of non-medallists had used "performance-enhancing" drugs when in competition (Pearson 1994).

In contrast to Olympic athletes, the second most prevalent group that is believed to abuse AS is the bodybuilding and/or weightlifting population. Yesalis et al 1998, reported that 55% of elite power lifters admit to AS usage. While Tricker et al 1989, reported the same percentage among amateur competitive bodybuilders. These high numbers relate to the fact that many bodybuilding competitions do not test for AS usage among its competitors. Although these reports seem high in contrast to the general population, our experience with competitive bodybuilders suggests these are underestimated. We have observed that since AS were categorized by the FDA as Schedule III class drugs of abuse, AS users have become far less forthcoming about their use of these drugs. This makes identification of AS using athletes by the general practitioner problematic.

Continuously increasing numbers of athletes are now relying upon anabolic steroids to enhance their strength, endurance and performance, despite the knowledge of the potentially serious adverse consequences these agents can have. Dezelsky et al (1985) noted an increase in anabolic steroids use. In 1970, 15% of all intercollegiate athletes used steroids. However, the figure increased to 20% in subsequent surveys conducted in 1976, 1980, and 1984. They also estimated that only 1% of non-athletic univer-sity students reported using steroids. These figures seem conservative when compared to more recent data on junior and senior high school students. In 1986, the Hazeldon Foundation of Minneapolis calculated the rate of previous or current use of anabolic steroids as being 3%, i.e., 5% for male and 1% for female students (Newman 1986). Buckley et al (1988) estimated that 6.6% of 12th grade male students have used or are still using steroids. Even more disturbing is their finding that 4.4% of male high school seniors initiated steroid use at 16 years of age or younger. More recently, it has been reported that 4-12% of US high school boys and 0.5-2% of US high school girls admit to using anabolic steroids at some time in their life (Yesalis and Bahrke 1995).

The anabolic steroid using athletes of today have a "sophisticated" pharmacologic knowledge, based on both their subjective experiences and anecdotal information, that surpasses the vast majority of physicians. For this reason, traditional warnings regarding the lack of efficacy and potential dangers of steroid abuse are universally held in con-tempt. Today it appears that the experts on anabolic steroids use in athletic competition are not medical clinicians but the athletes who for years have utilized themselves as the experimental subjects for high dose anabolic steroid usage. The time has come where the medical profession must recognize that scientifically relevant studies investigating the efficacy and toxicology of anabolic steroids are necessary. No longer is it acceptable to think it unethical to administer large doses of anabolic steroids in controlled double blind studies to athletes when there is a world-wide epidemic of anabolic steroid use among athletes at all levels of competition.

Perry et al (1990) conducted a study that chronicled the drug histories of a series of 20 competitive and noncompetitive weight lifters. Steroid efficacy studies only examine the anabolic effects of individual drugs. However, these 20 steroid users consistently practiced polypharmacy. During steroid use cycles lasting between 7-14 weeks, athletes would commonly utilize two or three oral agents and two long-acting injectable products. Dosages of oral preparations tended to be similar to those utilized in efficacy studies whereas dosages of the long-acting injectable agents were approximately three to eight times greater than those utilized in controlled studies. Usage onset in this group of subjects averaged 18.5 years old. Based on history, subjects reported significant increases in body weight and strength. Eighteen subjects reported a mean weight gain from 172 to 219 lbs (p < 0.0001). Strength as measured by the maximum bench press increased from a baseline of 257 to 370 lbs (p < 0.0001). This 47% increase in strength resulted from an average of 5.3 cycles of anabolic steroid use. Although these weight and strength gains are only based on the cohorts anecdotal reports, it is obvious that these results can be potentially addicting in individuals dissatisfied with their physical appearance.

Based on the pattern of anabolic steroid usage currently being practiced in the United States, it is apparent that past efficacy and toxicology studies are of limited value in delineating the benefits and hazards of these drugs.

Presentation

Case reports of mental status changes occurring secondary to the use or abuse of anabolic steroids began appearing in the medical literature in 1980 (Annitto and Layman 1980, Brower et al 1989, Pope and Katz 1987, Tennant et al 1988, Tilzey et al 1981). Annitto and Layman (1980) described a 17 year-old weight lifter that became depressed on two separate occasions while taking anabolic steroids. This individual also reported symptoms of suspiciousness and auditory hallucinations. Tilzey et al (1981) reported a case of delirium in a 66 year-old male while being treated with oxymethalone, 200-300 mg/d for anemia. Pope and Katz (1987) described two cases of mental disturbances potentially induced by anabolic steroids. Two patients experienced psychotic depressions, the first occurred within 14 days of the start of methyltestosterone 10 mg po bid while the second occurred after a second 8-week course of methandrostenolone, 15 mg po qd. Tennant et al (1988) described an anabolic steroid withdrawal reaction manifested by fatigue and depression. While taking the drugs, the patient complained of feelings of uncontrollable violence, paranoia, and suicidal ideation. Brower et al (1989) reported the first case of anabolic steroid dependence that met the DSM-III-R criteria for psychoactive substance. Tolerance, withdrawal symptoms, and the use of steroids to reverse withdrawal symptoms occurred. Thus symptoms of intoxication, withdrawal, and addiction have been anecdotally reported for the anabolic steroids. It could be argued that these reactions were simply coincidental to the steroid use.

Pope and Katz (1988) generated data that suggested these anecdotal reports were not the result of coincidental findings. Structured psychiatric interviews of 41 body-builders and football players who were steroid users found that 12% of the subjects met DSM-III-R criteria for mania while taking steroids and 12% for major depression when withdrawing from steroids. However, the authors did not utilize a control group of non-steroid using weight lifters and the findings may have been due to a group membership effect. However, Perry et al (1990) retrospectively evaluated the possible association of anabolic steroid abuse and mental status changes as well as the association of steroid use with major mental disorders by utilizing a control and a user group. The investigation characterized the symptom patterns and mental status changes precipitated by anabolic steroid abuse. Twenty male weight lifters that had used or were currently using anabolic steroids were compared to 20 male weight lifters that had never used steroids. Using the Symptom Checklist 90, the steroid users were found to have significantly (p < 0.005) more somatic, depressive, anxiety, hostility, and paranoid complaints when using steroids than when they were not using the drugs. When contrasted to the weight lifter controls, the steroid users had a significantly (p < 0.005) greater number of complaints of depression, anxiety, and hostility during cycles of steroid use. However, when utilizing the Diagnostic Interview Schedule to evaluate patients for the presence or occurrence of DSM-III diagnoses, no differences in the frequency of major mental disorders were found between the two groups. The authors concluded that the organic affective changes associated with anabolic steroid abuse usually present as a subsyndromal depressive disorder of insufficient severity to be classified as a psychiatric disorder.

Pope et al (1994) assessed the prevalence of psychiatric syndromes in a volunteer sample of 156 community athletes comprised of 88 steroid-users and 68 non-users. Subjects were administered the SCID to diagnose DSM-III-R syndromes in association with steroid use and in the absence of steroid use. Of the steroid users, 23% reported major mood syndromes--mania, hypomania, or major depression--in association with steroid use. Steroid users displayed mood disorders significantly more frequently during steroid exposure than in the absence of steroid exposure significantly and more frequently than nonusers. Users rarely abused other drugs simultaneously with steroids. Like the corticosteroids, the prevalence of steroid-associated major mood syndromes was dose dependent as shown in Table 1.

Table 1. Relationship of weekly steroid dose to prevalence of mood disorders.
	mania	hypomania	MDD	Total
low* (n=12)	0	1 (8)	0	1 (8)
medium* (n=51)	0	5 (10)	3 (6)	7 (14)
high* (n=25)	4 (16)	3 (12)	7 (28)	11 (44)
* Low = estimated < 300 mg/wk; medium = estimated 300-1000 mg/wk; high = estimated > 1000 mg/wk;

Utilizing the same weight lifter population described above, Yates et al (1990) conducted a personality disorder study. Because of the phenomenologic similarities between illicit anabolic steroid abuse and alcohol/drug abuse, they hypothesized that there might be an increased risk for antisocial personality disorder and possibly other personality disorder. Using the Personality Disorder Questionnaire (PDQ) as a test instrument, the steroid-users were first found to have an increased risk for personality psychopathology compared to an age- and sex-matched community control group. The risk appeared to be partially explained by a group membership effect. Secondly, when compared to a control group of alcoholics, illicit anabolic steroid users had a similar risk for having significant antisocial traits. The study has important clinical implications. If antisocial personality increases risk for anabolic steroid use as it does for alcohol and drug abuse, then prevention and treatment of steroid use may be quite complicated. Legal and interpersonal problems are likely to be found in many steroid abusers. Antisocial personality disorder complicating drug abuse is a negative predictor of psychotherapy outcome (Woody et al 1985). Motivation for treatment may be limited in the antisocial group, with limited response to the usual methods of substance abuse treatment.

Since antisocial personality often includes interpersonal aggression and vandalism, these individuals may be especially sensitive to the aggressive effects and complications of anabolic steroid use. In fact, previous animal studies have suggested the increase in aggression noted with anabolic steroid use is related to baseline behavior tendencies towards aggression (Rejeski et al 1988). Yates et al (1992) contrasted aggression/hostility in 12 current and past AS users and 24 non-AS using weight-lifters (control), with the Buss-Durke Hostility Inventory (BDHI). The steroid users' aggression factor items of direct assault, indirect assault, and verbal assault were higher than the controls. Importantly, the degree of aggression observed in the AS users, represented by the mean BDHI, was clinically significant and was higher than both a group of psychiatric patients and a group of prison inmates administered the BDHI scale.

AS and Aggression in healthy normal males
Although the presence of an underlying aggressive personality is a significant contributor to the increased aggressive tendencies reported with AS use, there is some indication that AS use can increase aggression in patients considered psychologically normal. Kouri et al (1995) examined the effect of testosterone cypionate (TC) on aggression in 6 individuals without an axis I diagnosis or ASPD. Three subjects had a past history of AS use, but were not current users (documented by urine screen). This was a randomized double-blind cross-over of testosterone versus placebo for 6 weeks each, with a 6 week treatment free period in between. Testosterone was administered as testosterone cypionate injections, 150 mg/week x 2 weeks, 300 mg/week x 2 weeks and then 600 mg/week x 2 weeks. The Point Subtraction Aggression Paradigm (PSAP), a computer simulated game that quantifies aggressive behavior towards a mock opponent, was performed at baseline and at the end of each treatment period. Testosterone administration was associated with higher levels of aggressive responses by this measure. This study successfully utilized a manipulation of the environment to provoke aggressive responding a strategy may be more sensitive than self-report measures of aggression. This experimental observation is consistent with anecdotal reports of aggressive outbursts in steroid abusers lacking significant prior psychiatric history. Pope and Katz (1990) described three male weight lifters that developed an organic mood disorder, manic episode while using steroids. All three impulsively committed a felonious act that lead to their incarceration, i.e., attempted murder and kidnapping, attempted murder, and murder. Notably, none of these individuals had a history of mental illness, including personality disorders, or aggressive or violent behavior.

Pope et al (2000) used the same crossover testosterone dosing strategy described above by Kouri et al (1995) in a larger sample of 56 males aged 20 to 50 years. In order to test a psychiatrically "normal" population, subjects were excluded if they had a current or past history of any major affective or psychotic illness or met DSM-III-R criteria for substance abuse during the past year. Significantly, Young Mania Rating Scale (YMRS) scores and PSAP scores were significantly higher (i.e. more impaired) while subjects were receiving testosterone compared to placebo. However, the distribution of these scores is also important. On average, the YMRS scores were 3 points higher in subjects while receiving testosterone. However, this mean difference does not indicate that all subjects had a 3 point increase, but rather that most subjects experienced no changes with testosterone and a few subjects had marked changes. When YMRS scores were categorized as marked (³ 20), moderate (10-19) and minimal (< 10), only one subject had a moderate score while on placebo with the remaining 49 evaluable subjects having minimal scores. While receiving testosterone, however, 42 subjects had minimal scores, 6 had moderate scores and 2 had marked YMRS scores.

Tricker et al (1996) also investigated the effects of testosterone administration to subjects without underlying psychopathology. The investigators randomized 53 eugonadal men, aged 19-40 years, to receive testosterone 600 mg/wk or placebo for 10 weeks. Each subject was also randomized to perform thrice-weekly exercise or no exercise. The primary outcome measures included the multidimensional anger inventory (MAI), the mood inventory (MI) and the observer mood inventory (OMI). Over the 10-week treatment period there were no significant differences in MAI, MI, or OMI scores between the testosterone and placebo groups. The authors concluded that doses of up to 600 mg/wk of testosterone do not increase angry behavior in normal males.

Perry et al (2003) attempted to characterize the relationship of aggression and testosterone blood levels among weightlifters who were users (n = 10) and nonusers (n = 18) of anabolic steroids. Participants were interviewed using the Modified Mania Rating Scale and Hamilton Rating Scale for Depression to assess mood, the Buss-Durkee Hostility Inventory (BDHI) and Point Subtraction Aggression Paradigm (PSAP) to assess aggression, and the Personality Disorder Questionnaire (PDQ-R) to assess personality. Blood samples were obtained for the determination of total, free, and weakly bound testosterone. Comparisons of continuous variables between testosterone users and non-users were performed with a parametric (unpaired t-test) or non-parametric (Mann-Whitney) test where appropriate. Correlations with testosterone were examined separately for testosterone users and non-users, using Spearman rank correlation. The subjective (BDHI) and objective (PSAP) assessments of aggression found that supranormal testosterone concentrations were associated with increased aggression. However, the PDQ-R results suggest that this finding was confounded by the personality disorder profile of the steroid users, because steroid users demonstrated Cluster B personality disorder traits for antisocial, borderline, and histrionic personality disorder.

In conclusion, studies involving the administration of testosterone to healthy normal men suggest that the majority of individuals will not experience psychiatric changes with doses up to 500-600 mg/wk. However, this response is not uniform and individual patients will experience marked affective changes, particularly as the dose increases beyond 500 mg/wk. Patients with underlying psychopathology (e.g. antisocial personality disorder) or a general predisposition towards anger are probably more likely to experience an increase in angry or aggressive behavior. This relationship is important because in practice these individuals are more likely to use anabolic steroids illicitly, compared to "healthy, psychologically normal men".

Psychosexual effects
The psychosexual effects of testosterone have also been studied (Yates et al 1999). The primary outcome measures were a daily rating of sexual interest and a diary of sexual activity. Other psychiatric assessment included the Buss-Durkee Hostility Inventory, the BPRS, the Modified Mania Rating Scale and the HAMD. Following two initial weeks of single-blinded placebo injections, subjects received injections of either 100mg, 250 or 500mg of testosterone cypionate per week for 14 weeks. Overall, the subjects had no significant changes in sexual interest or activity. Additionally, there were no clinically significant changes in any of the other psychological assessments. However, one patient was withdrawn as a result of significant hypomanic personality changes. The authors concluded that supraphysiologic testosterone administration had a minimal psychosexual risk overall, but some patient may experience significant psychiatric changes. The authors also noted that the doses utilized in this study might be lower than taken by many testosterone abusers.

Table 2. Self-reported symptoms of dependence (n=49).
DSM-III-R criteria	Number (%)
More AS use than intended	25 (51)
Desire yet unable to cut down to control use	8 (16)
Large time expenditure on substance-related activity	19 (40)
Frequent intoxication or withdrawal symptoms when expected to function or when physically hazardous	4 (9)
Social, work, or leisure activities replaced by AS use	14 (29)
Continued AS use despite problems caused or worsened by use	18 (37)
Tolerance	9 (37)
Withdrawal symptoms	41 (84)
Substance used to relieve or avoid withdrawal symptoms	2 (4)

Muscle Dysmorphia
Olivardia et al (2000) conducted a case-control comparison trial of 24 male weight-lifters with muscle dysmorphia and 30 normal comparison weight-lifters. Currently muscle dysmorphia is not a DSM-IV recognized diagnosis. Pope has suggested that muscle dysmorphia is an unrecognized form of body dysmorphic disorder (Pope et al 1997). Table 3 presents the responses of the two groups to the muscle dysmoprhia questionnaire.

Table 3. Response on muscle dysmorphia symptom questionnaire of men with muscle dysmorphia vs normal controls respectively (Olivardia et al 2000).
Question	subjects vs controls (Number or %)
How often do you weigh yourself per week?	5 vs 2*
How often do you check mirrors in a day	9 vs 3*
How many minutes in a day are you preoccupied by being too small, not being big enough, or getting bigger?	325 vs 41*
Have you ever worn heavy sweatshirts in the summer or refused to take your shirt off in public for fear that someone may think you were too small?	88% vs 0%*
Have you ever given up enjoyable activities to go to the gym to get bigger?	100% vs 37%*
*p < 0.05

Morbidity and Mortality
Thiblin et al (2000) described a series of medicolegally investigated deaths of 34 white male (20-45 yo) users of anabolic steroids. The causes of death were homicide (n=11), suicide (n=9), accidents (n=12) and indeterminate (n=2). Chronic myocardial pathological changes, e.g., myocardial hypertrophy, coronary arteriosclerosis, myocardial fibrosis, were observed in 12 of the 34 deaths. The homicides, suicides, and accidental drug overdoses were related to impulsive and disinhibited behavior marked by violent rages, mood swings, and uncontrolled drug intake e.g, ethanol, cocaine/amphetamines, opiates, and benzodiazepines. Only 5 of the 33 urine drug abuse screens were negative.

References

Annitto WJ, Layman WA (1980). Anabolic steroids and acute schizophrenic episode. J Clin Psychiatry 41:143-4.

Boston Collaborative Drug Surveillance Program (1972). Acute adverse reactions to prednisone in relation to dosage. Clin Pharmacol Therap 13:694-8.

Brower KJ, Blow FC, Beresford TP, et al (1989). Anabolic-androgenic steroid dependence. J Clin Psychiatry 50:31-3.

Brower KJ, Blow FC, Young JP, Hill EM (1991). Symptoms and correlates of anabolic-androgenic steroid dependence. Br J Addiction 86;759-68.

Buckley WE, Yesalis CE, Friedl KE, et al (1988) Estimated prevalence of anabolic steroid use among male high school seniors. JAMA 260:3441-3445.

Bunim JJ, Ziff M, McEwen C (1955). Evaluation of prolonged cortisone therapy in rheumatoid arthritis--a four year study. Am J Med 18:27-40.

Cade R, Spooner G, Schlein E, et al (1973). Comparison of azathioprine, prednisone, and heparin alone or combined in treated lupus nephritis. Nephron 10:37-56.

Choi PYL, Parrott AC, Cowan D (1990). High-dose anabolic steroids in strength athletes: effect upon hostility and aggression. Human Psychopharmacol 5:439-56.

Clark LD, Bauer W, Cobb S (1952). Preliminary observations on mental disturbances occurring in patients under therapy with cortisone and ACTH. New Eng J Med 246:205-16.

Clark LD, Quarton GC, Cobb S (1953). Preliminary observations on mental disturbances occurring in patients under therapy with cortisone and ACTH Therapy. New Eng J Med 249:178-83.

Dezelsky TL, Toohey JV, Shaw RS (1985). Non-medical drug use behavior at five United States universities: a 15-year study. Bull Narc 37:49-53.

Falk WE, Mahnke MW, Peskanzer DC (1979). Lithium prophylaxis of corticotrophin- induced psychosis. JAMA 241:1011-12.

Hall RC, Popkin MK, Stickney SK, et al (1979). Presentation of steroid psychosis. J Nerv Ment Dis 167:229-36.

Haskett RF (1985). Diagnostic categorization of psychiatric disturbance in Cushing's syndrome. Am J Psychiatry 142:911-6.

Hayreh SS, Watson PG (1970). Prednisolone-21-stearoylglycolate in scleritis. Brit J Ophthalmol 54:394-8.

Kimball CP (1971). Psychological dependency on steroids? Ann Int Med 75:111-13.

Kreus KE, Viljanen AA, Kujale E, et al (1975). Treatment of steroid-dependent asthma patients with beclomethasone dipropionate aerosol. Scand J Resp Dis 56:47-57.

Kouri EM, Lukas SE, Poper HG, et al (1995). Increased aggressive responding in male volunteers following the administration of gradually increasing doses of testosterone cypionate. Drug Alcohol Depend 40:73-9.

Lewis D, Smith R (1983). Steroid-induced psychiatric syndromes: a report of 14 cases and a review of the literature. J Affective Disord 5:319-32.

Ling MH, Perry PJ, Tsuang MT (1981). Side effects of corticosteroid therapy: psychiatric aspects. Arch Gen Psychiatry 38:741-7.

Marx FW, Barker WF (1967). Surgical results in patients with ulcerative colitis. Am J Surg 113:157-64.

McCawley A (1965). Cortisone habituation-a clinical note. New Eng J Med 273:976.

Morgan HG, Boulnois J, Burns-Cox C (1973). Addiction to prednisone. Br J Med 2:93-4.

Newman M (1986). Michigan Consortium of Schools Student Survey. Minneapolis, Hazelden Research Services.

Nielsen JB, Drivsholm AA, Fischer F, et al (1963). Long-term treatment with cor-ticosteroids in rheumatoid arthritis. Acta Med Scand 173:177-83.

Olivardia R, Pope HG, Hudson JI (2000). Muscle dysmorphia in male weight-lifters: a case-control study. Am J Psychiatry 157:1291-6.

Pearson B. Olympic survey: Olympians of winters past. USA Today 1994; Feb 7th:C5

Perry PJ, Tsuang MT, Hwang MH (1984). Prednisolone psychosis: clinical observations. Drug Intel Clin Pharm 18:603-8.

Perry PJ, Andersen KH, Yates WR (1990a). Illicit anabolic steroid use in athletes: a case series analysis. Am J Sports Med 18:420-8.

Perry PJ, Yates WR, Andersen KH (1990b). Psychiatric effects of anabolic steroids: a controlled retrospective study. Ann Clin Psychiatry 2:11-7.

Perry PJ, Kutscher EC, Lund BC, et al (2003). Measures of aggression and mood changes in male weightlifters with and without androgenic anabolic steroid use. J Forensic Sci 48:646-51.

Pope HG, Katz DL (1987). Bodybuilder's psychosis. Lancet 1:863.

Pope HG, Katz DL (1988). Affective and psychotic symptoms associated with anabolic steroid use. Am J Psychiatry 145:487-90.

Pope HG, Katz DL (1990). Homicide and near-homicide by anabolic steroid users. J Clin Psychiatry 51:28-31.

Pope HG, Katz DL (1994). Psychiatric and medical effects of anabolic-androgenic steroid use. A controlled study of 160 athletes. Arch Gen Psychiatry 51:375-82.

Pope HG, Kouri EM, Hudson JI (2000). Effects of supraphysiologic doses of testosterone on mood and aggression in normal men. Arch Gen Psychiatry 57:133-40.

Pope HG, Gruber AJ, Choi P, et al (1997). "Muscle dysmorphia": an underrecognized form of body dysmorphia disorder? Psychosomatics 38:548-57.

Rejeski WJ, Brubaker PH, Herb RA, et al (1988). Anabolic steroids and aggressive behavior in cynomolgus monkeys. J Beh Med 11:95-105.

Rosenberg FR, Sanders S, Nelson CT, et al (1976). Pemphigus. Arch Dermatol 112:962-70.

Sergent JS, Lockshin MD, Klempner MS, et al (1975). Central nervous system disease in systemic lupus erythematosus. Am J Med 58:644-54.

Smyllie HC, Connolly CK (1968). Incidence of serious complications of corticosteroid therapy in respiratory disease. Thorax 23:571-81.

Sutor B, Wells LA, Rummans TA (1996). Steroid-induced depressive psychosis responsive to electroconvulsive therapy. Convulsive Therapy 12:104-7.

Sullivan ML, Martinez CM, Gennis P et al (1998). The cardiac toxicity of anabolic steroids. Prog Cardiovasc Dis 41:1-15.

Tennant F, Black DL, Voy RO (1988). Anabolic steroid dependence with opioid- type features. N Eng J Med 1319:578.

Thiblin Im Lindquist O, Raja J (2000). Cause and manner of death among users of anabolic androgenic steroids. J Forensic Sci 45:16-23.

Tilzey A, Heptonstall J, Hamblin T (1981). Toxic confusional state and choreiform movements after treatment with anabolic steroids. Br Med J 283:349-50.

Tricker R, O'Neill M, Cook D (1989). The incidence of anabolic steroid use among competitive bodybuilders. J Drug Educ 19:313-25.

Tricker R, Casaburi R, Storer TW et al (1996). The effects of supraphysiological doses of testosterone on angry behavior in healthy eugonadal men: a clinical research center study. J Clin Endocrinol Metab 81:3754-8.

Wolkowitz OM (1994). Prospective controlled studies of the behavioral and biological effects of exogenous corticosteroids. Psychoneuroendocrinol 19:233-55.

Wolkowitz OM and Reus VI (1999). Treatment of depression with antiglucocorticoid drugs. Psychosom Med 61:698-711.

Wolkowitz OM, Reus VI, Chan T, et al (1999). Antiglucocorticoid treatment of depression: double-blind ketoconazole. Biol Psychiatry 45:1070-4.

Woody GE, McLellan T, Luborsky L, et al (1985). Sociopathy and treatment outcome. Arch Gen Psychiatry 42:1081-6.

Yates WR, Perry PJ, Andersen KH (1990). Illicit anabolic steroid use: a controlled personality study. Acta Psychiatr Scand 81:548-50.

Yates WR, Perry PJ (1992). Aggression and hostility in anabolic steroid users. Biol Psychiatry 31:1232-4. The

Yates WR, Perry PJ, MacIndoe J, et al (1999). Psychosexual effects of three doses of testosterone cycling in normal men. Biol Psychiatry 45:254-60.

Yesalis CE, Herrick R, Buckley W, et al (1988). Self-reported use of anabolic-androgenic steroids by elite power lifters. Physician Sportmed 16:91-100.

Yesalis CE, Bahrke MS (1995). Anabolic-androgenic steroids: current issues. Sports Med 19:326-40.

Yesalis CE, Buckley WE, Anderson WA, et al (1990). Athletes' projections of anabolic steroid use. Clin Sports Med 2:155-71.

Title Page