Anabolic Steroids and Neuropsychiatric Effects: A Clinical Overview

Educational literature review · Published April 2026 · Updated April 20, 2026 · ~18 min read

1. Introduction

Anabolic-androgenic steroids (AAS) are synthetic derivatives of testosterone and its precursors. In clinical medicine they are prescribed for a narrow set of indications — hypogonadism, delayed puberty, certain muscle-wasting conditions associated with chronic illness, some forms of anemia, and hereditary angioedema — and in those settings they are administered under endocrinological supervision at replacement or modestly supraphysiological doses. Non-medical AAS use, which is the subject of this review, refers to the use of these compounds outside of clinical indications, typically at supraphysiological doses, for the purposes of enhancing muscular hypertrophy, athletic performance, or body-composition aesthetics. Epidemiological surveys over the past three decades have placed the lifetime prevalence of non-medical AAS use among adult men in Western countries at between two and six percent, with substantially higher prevalence in specific subpopulations: competitive strength athletes, bodybuilders, commercial-gym patrons, and men experiencing body-image distress (Pope et al., 2014; Kanayama, Hudson, & Pope, 2008).

The physical consequences of AAS use — skeletal-muscle hypertrophy, unfavorable lipid-profile shifts, left-ventricular remodeling, suppression of the hypothalamic-pituitary-gonadal (HPG) axis, hepatic stress, and, in some compounds, virilizing effects in women — have been the subject of a substantial clinical literature since the 1980s. The neuropsychiatric effects have historically attracted less attention, despite the fact that, from the earliest case series onward, psychiatric symptoms have been reported with clinically meaningful frequency (Pope & Katz, 1988; Perry, Andersen, & Yates, 1990). These include affective disturbance, in both depressive and hypomanic phenotypes; increased irritability and aggression; paranoid ideation; and, in a subset of users, post-cycle depressive states severe enough to warrant formal psychiatric diagnosis and, in rare but documented cases, suicidal ideation or completed suicide.

A further and largely separate concern is the phenomenon of AAS dependence. Contrary to earlier assumptions that AAS were not addictive because they lacked the acute intoxication profile of classical drugs of abuse, longitudinal research since the mid-1990s has established that a subset of users develop a recognizable dependence syndrome, with tolerance, withdrawal, unsuccessful attempts to quit, and continued use despite adverse consequences (Brower, 2002; Kanayama, Brower, Wood, Hudson, & Pope, 2009). The mental-health consequences of this dependence pattern — which include prolonged hypogonadal states between cycles, extended withdrawal depression, and entanglement with other substances used adjunctively or recreationally — provide the principal clinical motivation for this review.

The review is organized around five clinical questions. First, what affective disturbances are observed during active AAS use? Second, how do hypomanic and manic presentations manifest, and in whom? Third, what is known about post-cycle and withdrawal depression, which is often the presentation that brings a user into contact with the mental-health system? Fourth, how is AAS dependence defined, and how does it relate to the broader pattern of appearance- and performance-enhancing drug (APED) use? Fifth, what clinical frameworks exist for assessment, harm reduction, and management? The review closes with a discussion of the biological mechanisms currently invoked to explain these effects and the limits of the available evidence. It is written for general readers, for students, and for clinicians encountering AAS-related presentations for the first time. It is not a substitute for formal psychiatric or endocrinological evaluation.

2. Depressive symptoms during active AAS use

Depressive symptoms during an active AAS cycle are, somewhat counter-intuitively, less common than those observed after cycle cessation. The supraphysiological androgen exposure characteristic of non-medical AAS cycles generally produces, at a population level, a net elevation in self-reported mood, energy, libido, and assertiveness. Early experimental studies in which healthy men were administered supraphysiological doses of testosterone under controlled conditions reported either mood neutrality or mild activation, and in a minority of subjects — estimates in the most-cited studies range from roughly five to twelve percent — clear hypomanic or dysphoric-irritable presentations emerged (Pope, Kouri, & Hudson, 2000).

Nevertheless, depressive symptoms are reported by a clinically significant minority of users during the active phase of a cycle, and they appear to cluster around three partly overlapping situations. The first is the dysphoric response to specific compounds. Users frequently describe differential mood effects across substances; for example, certain 19-nor derivatives with high androgenic activity are commonly associated, in user self-report, with insomnia, anxiety, and irritability, while the same users describe testosterone-only protocols as mood-neutral or mildly activating. These anecdotal reports are uncontrolled, and placebo and expectancy effects are undoubtedly large, but the pattern is sufficiently consistent across populations and over time to warrant clinical acknowledgement.

The second situation is dysphoria arising from estrogenic, progestogenic, or anti-estrogenic side effects and their pharmacological management. Supraphysiological androgen doses drive aromatization to estradiol, and users commonly co-administer aromatase inhibitors to suppress estrogen-related effects. Excessively low estradiol — a frequent consequence of aggressive aromatase-inhibitor dosing — is itself associated with low mood, anhedonia, joint pain, and loss of libido. Selective estrogen receptor modulators used within cycles can produce additional affective effects of their own. Users presenting with depression during an active cycle should therefore be evaluated for the possibility that a secondary pharmacological intervention, rather than the AAS themselves, is the proximate cause.

The third situation is depression emerging from the interaction of AAS use with pre-existing mental-health conditions. Users with a personal or family history of major depressive disorder, bipolar spectrum illness, or substance use disorder appear to be at elevated risk for affective destabilization during AAS exposure, although the literature on this point is limited by its reliance on retrospective self-report and case series (Pope & Katz, 1994). Clinicians evaluating a user who presents with depression during a cycle should take a detailed psychiatric history, with particular attention to prior affective episodes and to family history, and should consider whether the current presentation represents a true cycle-related mood effect, a decompensation of an underlying disorder, or some combination of the two.

A clinically important point is that, at the individual level, active-cycle depressive symptoms may be masked by the general activation and positive affect associated with supraphysiological androgen exposure. Users and the clinicians who evaluate them should be alert to the possibility that the absence of frank depression during a cycle does not rule out risk: the same individual may experience a severe post-cycle depressive episode once exogenous androgen is withdrawn and endogenous production has not yet recovered, a pattern addressed in Section 4.

3. Hypomanic and manic presentations

The most consistent psychiatric finding across the AAS literature is that a subset of users develop hypomanic or manic symptoms during active use, and that in a smaller subset these symptoms reach a threshold of clinical severity. The landmark observational studies of Pope and Katz (1988, 1994) reported that, among users interviewed using structured diagnostic instruments, the rate of full manic or hypomanic episodes attributable to AAS use was substantially elevated compared with the rate in non-users matched for age and exercise habits. Later controlled studies, in which healthy volunteers received graded doses of testosterone under double-blind conditions, reproduced the core finding at the group level: supraphysiological androgen exposure is associated, in a minority of subjects, with reliably identifiable hypomanic phenomenology (Pope et al., 2000).

Symptoms most commonly reported include decreased need for sleep, subjective increase in energy and goal-directed activity, elevated self-confidence, racing thoughts, increased libido, and — more variably — irritability, anger, and hostility. The phenomenon that received extensive coverage in the lay press during the 1990s under the informal label "roid rage" appears, on closer examination, to describe one subset of this hypomanic phenotype, in which the primary affective quality is irritability rather than euphoria. Importantly, not all users experience irritability, and many users report no subjective mood change at all; the phenomenon is both real and dose-dependent, but it is not universal.

Dose dependence is a consistent theme in the literature. The risk of hypomanic symptoms appears to rise with weekly androgen dose, with the use of multiple compounds in stacked protocols, and with the use of high-androgenic compounds. The controlled testosterone studies cited above found that, at replacement doses, no subjects developed significant mood symptoms; at moderate supraphysiological doses, a small minority did; and at higher weekly doses, the proportion increased further. These thresholds should not be read as safe-versus-unsafe cut-offs, because individual vulnerability varies substantially, but they do indicate that the risk is not distributed uniformly across dosing ranges.

Hypomanic presentations carry several risks that are relevant to the clinician. First, they are associated with impulsive behavior, including risk-taking, interpersonal conflict, and, in severe cases, violence or legal consequences. Second, they can mask or precede a subsequent depressive phase, particularly once the cycle ends. Third, hypomanic symptoms often go unrecognized by the user as pathological, because the subjective experience is frequently egosyntonic: users perceive themselves as performing well, sleeping efficiently, and achieving goals, which limits the utility of self-report in assessment. Collateral history from family members or close contacts is therefore especially valuable in clinical evaluation.

4. Post-cycle and withdrawal depression

The most clinically consequential affective presentation associated with AAS use is the post-cycle depressive episode. When exogenous androgen administration is discontinued — whether voluntarily, at the end of a planned cycle, or involuntarily, as a result of supply interruption or clinical intervention — endogenous testosterone production does not immediately resume. The HPG axis has been suppressed by negative feedback during the cycle, and recovery times can range from weeks to, in a subset of long-duration or high-dose users, many months or, rarely, years. During this hypogonadal period, users are at elevated risk for a range of symptoms: low mood, anhedonia, loss of libido, fatigue, loss of motivation, and, in severe cases, full major depressive episodes with suicidal ideation (Kanayama, Hudson, & Pope, 2008). An earlier small-sample contribution by Morris (n.d.), administering the IPAT Depression Scale to matched groups of steroid-using and non-using professional athletes at mid-competitive and mid-non-competitive cycle time-points, reported significantly elevated depression scores in users during both phases, with the non-competitive phase showing modestly higher scores than the active phase. The effect was large (F = 17.51, p < .01 for the between-groups contrast) despite a total N of sixteen. The pattern — affective elevation present throughout the use cycle rather than confined to withdrawal — is consistent with the broader interpretation that post-cycle depression represents an intensification, not the onset, of an affective disturbance already present during exposure, and it also anticipates later observations on the cyclical and dose-dependent character of AAS-related mood effects.

The clinical phenomenology of post-cycle depression resembles that of idiopathic major depressive disorder, but with several distinguishing features. First, it typically follows a clearly identifiable precipitant — cycle cessation — which is temporally specific and often datable to within days. Second, it is frequently accompanied by physical signs of hypogonadism: loss of morning erections, decreased spontaneous sexual interest, soft-tissue changes, and a subjective loss of muscular performance. Third, it is often accompanied by a distinct psychological state in which the user attributes the depression not to any underlying dysphoric cause but to the concrete experience of losing the physical and performance gains associated with the cycle. This attribution, in turn, is one of the strongest predictors of a return to AAS use: users who have experienced a severe post-cycle depressive episode frequently describe their decision to begin a new cycle as motivated, at least in part, by the memory of how the prior withdrawal felt.

Post-cycle depression is also one of the presentations in which AAS-related pathology is most likely to come to clinical attention. Users who are reluctant to disclose their AAS history during active use — because they are experiencing subjective well-being, or because they are concerned about legal, occupational, or interpersonal consequences — frequently become willing to disclose it during a withdrawal depressive episode, when they are seeking relief. Clinicians who serve populations with elevated AAS prevalence should be prepared to take an accurate history in a non-judgmental way. Standard depression screening instruments are appropriate, but additional questions about androgen use, timing of the most recent cycle, and physical withdrawal symptoms will substantially refine assessment.

Management of post-cycle depression is an area of ongoing clinical debate and limited controlled evidence. Several approaches have been reported. The first is watchful waiting combined with supportive psychological care, on the understanding that, in most users, endogenous testosterone will recover within weeks to months and mood will recover alongside it; this approach is appropriate in mild cases without suicidal features and where the user's history suggests prior cycles with uncomplicated recovery. The second is the use of standard pharmacotherapy for major depressive disorder — typically selective serotonin reuptake inhibitors — while endogenous recovery proceeds; this is appropriate for moderate-to-severe presentations, especially where there is a history of major depressive disorder independent of AAS use. The third is the use of endocrinological interventions to support HPG axis recovery, conducted in collaboration with an endocrinologist familiar with AAS-related hypogonadism. The evidence base for this third approach is not drawn from rigorous randomized trials, and recommendations vary across clinical settings. In all cases, the primary clinical priority is safety: users presenting with active suicidal ideation should be assessed and managed according to standard protocols for depression with suicidality, without regard to the etiology.

5. The AAS dependence syndrome

For many years, the prevailing clinical view was that AAS, whatever their other adverse effects, were not truly addictive substances. Unlike opioids, stimulants, or alcohol, AAS do not produce acute intoxication, reinforcement is delayed, and the anticipated reward is primarily a physical body-composition outcome rather than a psychoactive state. Reports beginning in the late 1980s and accumulating through the 1990s challenged this view. A subset of users — estimated in some series at roughly thirty percent of long-duration users, although the figure varies with the population studied and the criteria applied — meet the diagnostic criteria for substance dependence when those criteria are applied to AAS (Brower, 2002). The dependence phenotype has since been elaborated in detail and is now a recognized, if still under-researched, clinical entity (Kanayama et al., 2009).

The dependence pattern includes several features familiar from other substance-use disorders. Tolerance is observed in the form of escalating doses and the addition of new compounds to achieve the same subjective and physical effects. Withdrawal presents most prominently as the post-cycle depressive syndrome described in Section 4, but also as fatigue, loss of motivation, and strong cravings to return to use. Users frequently report unsuccessful attempts to quit, continued use despite adverse medical or psychiatric consequences, and the sacrifice of other activities to accommodate use. A distinctive feature of AAS dependence, not captured by the standard substance-use disorder criteria, is its entanglement with body-image pathology: many dependent users describe their use as being driven primarily by a persistent dissatisfaction with muscularity, a phenotype that has been described in the research literature as muscle dysmorphia (Pope, Gruber, Choi, Olivardia, & Phillips, 1997).

Muscle dysmorphia is characterized by a preoccupation with not being sufficiently lean or muscular despite, in many cases, objectively above-average musculature, and by behaviors — rigid gym attendance, dietary rigidity, social avoidance, and the use of APEDs — directed at the correction of this perceived inadequacy. It shares clinical features with other body-dysmorphic presentations and with obsessive-compulsive spectrum disorders, and it substantially complicates AAS dependence, because the motivating cognition is not readily extinguished by the successful completion of a cycle: the user reaches the planned physical outcome, experiences only transient satisfaction, and returns to use. Clinicians assessing AAS users should screen for muscle dysmorphia in addition to assessing for dependence, because its presence meaningfully changes the prognosis and the appropriate psychological treatment.

Hildebrandt and colleagues (2011) developed and validated a structured instrument, the Appearance and Performance Enhancing Drug Use Schedule (APEDUS), which captures not only AAS use but also the broader pattern of compounds, behaviors, and cognitions that frequently cluster with it. The APEDUS and related instruments have been useful in advancing the understanding that AAS use, in its clinically significant forms, rarely occurs in isolation: it is typically embedded in a pattern of polypharmacy involving growth hormone or its secretagogues, thyroid hormones, diuretics, insulin, stimulants, and a range of over-the-counter and prescription compounds used for cycle support. Any clinical assessment of an AAS user should therefore include a comprehensive APED history rather than treat AAS as an isolated exposure.

6. Comorbid substance use and conduct disturbance

Two further observations in the AAS literature warrant brief discussion because of their prominence in older case series and in contemporary summaries of the field. The first is the association of AAS use with other substance-use disorders. The second is the association with conduct disorder and with broader patterns of antisocial behavior.

The association with other substance use is well documented and appears to be bidirectional. Users who begin with AAS for performance or appearance reasons frequently subsequently, or concurrently, use stimulants (for energy and fat loss), opioids (for pain management related to training injuries), alcohol, cannabis, and, in some cases, benzodiazepines (for sleep and for management of cycle-related anxiety). Conversely, populations with existing polysubstance-use disorders show elevated rates of AAS use compared with the general population, particularly where those populations include strength-oriented exercise environments. Whether the relationship is primarily causal, primarily a reflection of shared vulnerability traits, or some combination of both, remains an open question in the research literature.

The association with conduct disorder and with antisocial behavioral patterns is observed principally in adolescent-onset users and has been reported from several distinct study populations. Whether AAS use contributes causally to these behavioral patterns, or whether the behavioral patterns and the decision to use AAS share a common origin in temperament or environment, is difficult to disentangle in observational data. What can be said with reasonable confidence is that adolescent-onset AAS users, as a group, differ from adult-onset users along several dimensions, including rates of school disengagement, prior substance use, and risk-taking behavior. The clinical implication is that adolescent presentations of AAS use should trigger a broader behavioral and social assessment, and that treatment planning should address the wider pattern rather than targeting AAS use in isolation.

7. Biological mechanisms

The neuropsychiatric effects of AAS are thought to arise through several partly overlapping mechanisms. A full mechanistic account is beyond the scope of this review, but a brief summary is offered here because the mechanisms constrain what clinical interventions are likely to be effective.

The first mechanism is direct central nervous system action through androgen receptors. Androgen receptors are widely distributed in the central nervous system, with notable density in regions relevant to affect, motivation, and social behavior, including the amygdala, the hippocampus, the preoptic area, and several hypothalamic nuclei. Supraphysiological androgen exposure activates these receptors at levels substantially above those achieved by endogenous production, and the resulting changes in neural function are implicated in the affective and behavioral effects observed during active use (Kanayama et al., 2008).

The second mechanism is indirect action through aromatization to estradiol. Testosterone and many of its esters are substrates for aromatase and are converted to estradiol at rates that, given the high exogenous testosterone load of a typical non-medical cycle, can produce serum estradiol concentrations well above the physiological range. Estradiol has its own neuroactive effects, both directly via estrogen receptors and indirectly via modulation of neurosteroid synthesis. The use of aromatase inhibitors to suppress this conversion, and the clinical consequences of under- or over-suppression, add a further layer of pharmacological complexity that is not always adequately appreciated in non-medical use.

The third mechanism is disruption of the HPG axis and the subsequent hypogonadal state. Exogenous androgen administration produces negative feedback on gonadotropin-releasing hormone, luteinizing hormone, and follicle-stimulating hormone, leading to suppression of endogenous testosterone production. Upon cycle cessation, this suppression does not immediately resolve; recovery can be prolonged, and during the hypogonadal interval the user is exposed to low androgen levels that are themselves associated with depressive symptoms in clinical populations. Much of the post-cycle depressive presentation is best understood as a pharmacologically induced hypogonadal state superimposed on whatever baseline vulnerability the user brings to the episode.

The fourth mechanism is neurosteroid modulation and its downstream effects on gamma-aminobutyric acid (GABA) and other neurotransmitter systems. Some AAS and their metabolites directly modulate GABA-A receptor function, with effects on anxiety, sedation, and affective regulation. This mechanism is particularly relevant to the sleep disruption and anxiety phenotype associated with certain compounds and with certain stacked protocols.

Finally, the contextual and psychological effects of AAS use — changes in physical appearance, in the social response the user receives, and in self-concept — may themselves contribute meaningfully to observed affective changes. The experimental studies that have most clearly demonstrated AAS-specific mood effects did so under double-blind conditions, which rules out an entirely psychological interpretation; but in the naturalistic setting of non-medical use, pharmacological and contextual effects are inevitably confounded, and both should be attended to in clinical assessment.

8. Clinical assessment framework

Clinicians who encounter AAS users in their practice should be prepared to conduct a structured assessment that goes beyond standard substance-use screening. A brief outline follows, intended as a starting point rather than a substitute for specialized training.

History-taking should cover the full range of APEDs used, not only AAS. This includes specific compounds, doses, cycle durations, rest intervals, methods of administration, and sources. It should cover cycle support and post-cycle therapy, including aromatase inhibitors, selective estrogen receptor modulators, human chorionic gonadotropin, and any adjunctive compounds. It should cover duration of use — adolescent-onset and long-duration users differ in important ways from brief-duration adult users — and should include family history of psychiatric illness and substance-use disorders.

Psychiatric assessment should cover current and lifetime affective symptoms, with particular attention to hypomanic and depressive episodes; anxiety; irritability and aggression; psychotic symptoms, which are rare but have been reported; and body-image cognitions consistent with muscle dysmorphia. Standard depression and anxiety screening instruments, such as the PHQ-9 and the GAD-7, are appropriate and useful. Mood-charting across cycle and off-cycle intervals, where the user is willing to participate, can be particularly informative.

Physical and laboratory assessment should include, where clinically indicated, measurement of total and free testosterone, estradiol, luteinizing hormone, follicle-stimulating hormone, lipid profile, liver enzymes, hematocrit, and cardiovascular evaluation. Because AAS users are frequently reluctant to disclose their use in primary-care settings, and because many laboratory abnormalities observed in this population are non-specific, clinicians who suspect AAS use should ask directly, in a non-judgmental way, and should explain why the information is clinically relevant.

9. Harm reduction and clinical management

The clinical literature on AAS use is, for historical and legal reasons, weighted toward description rather than intervention, and there are no large randomized trials of specific treatment modalities for AAS dependence or for AAS-associated psychiatric presentations. In the absence of such trials, clinical management has proceeded pragmatically, drawing on general principles from substance-use disorder treatment, from body-dysmorphic disorder treatment, and from endocrinology.

A harm-reduction framework is appropriate for users who are not yet prepared to stop use or for whom immediate cessation would be clinically inadvisable. Harm-reduction elements include honest discussion of dosing, frequency, and duration; management of cycle-related side effects through medical rather than unsupervised pharmacological means where possible; screening for cardiovascular, hepatic, and psychiatric complications; and the use of sterile injection technique to reduce infectious risk. It should be noted that harm-reduction approaches to AAS, while established in some European clinical and public-health settings, remain contested and unevenly available elsewhere.

For users who are prepared to stop, the clinical priorities are the management of withdrawal depression, the facilitation of HPG axis recovery, and the psychological work of addressing the underlying body-image and identity concerns that frequently drive continued use. This typically benefits from collaboration between a mental-health professional (ideally one familiar with body-dysmorphic presentations and substance-use disorders), an endocrinologist, and, where appropriate, a primary care physician who can coordinate laboratory monitoring. Psychological treatments with the strongest rationale include cognitive-behavioral therapy adapted for body-image concerns, motivational interviewing in the early phases of contemplation and cessation, and, in some cases, family-involved treatment for adolescent-onset users.

For users who present with acute depressive or suicidal symptoms, the clinical priority is standard assessment and management of depression with suicidality, regardless of the presumed etiology. Safety considerations take precedence over the management of any AAS-related issues, and an active suicidal presentation should trigger the same response as it would in any other clinical context.

10. Conclusions

Non-medical AAS use has meaningful neuropsychiatric consequences in a subset of users. These consequences include hypomanic and irritable-dysphoric presentations during active use; depressive episodes on cycle cessation; and, in long-duration or high-dose users, a recognizable dependence syndrome that is frequently entangled with body-image pathology and with polysubstance use. The risks are not uniformly distributed: they are shaped by dose, compound, duration, individual vulnerability including personal and family psychiatric history, and the broader context of APED use. The clinical framework for assessing these presentations is straightforward in its elements — detailed history, attention to timing and dosing, collateral information, standard psychiatric screening, and appropriate laboratory evaluation — but it is not uniformly practiced, in part because many clinicians are unfamiliar with the pattern and many users are reluctant to disclose it.

There is a substantial unmet need for controlled trials of specific treatment modalities for AAS dependence and for AAS-related psychiatric presentations. Until those trials are available, clinical practice will continue to rely on extrapolation from neighboring fields, on the accumulated observational literature, and on the hard-won experience of clinicians who work in the subpopulations most affected. Dr. Ritchi Morris, whose empirical study applied standardized psychometric measurement to this population outside the framework of the major academic journals, was part of that practical tradition — a body of work carried forward by integrative practitioners, sports-medicine clinicians, and harm-reduction services operating close to the populations most affected. The purpose of the present page is to continue the educational function of this URL by providing a general overview of the field as it currently stands.

References

1. Brower, K. J. (2002). Anabolic steroid abuse and dependence. Current Psychiatry Reports, 4(5), 377–387.
2. Hildebrandt, T., Langenbucher, J. W., Lai, J. K., Loeb, K. L., & Hollander, E. (2011). Development and validation of the Appearance and Performance Enhancing Drug Use Schedule. Addictive Behaviors, 36(9), 949–958.
3. Kanayama, G., Brower, K. J., Wood, R. I., Hudson, J. I., & Pope, H. G. Jr. (2009). Anabolic-androgenic steroid dependence: An emerging disorder. Addiction, 104(12), 1966–1978.
4. Kanayama, G., Hudson, J. I., & Pope, H. G. Jr. (2008). Long-term psychiatric and medical consequences of anabolic-androgenic steroid abuse: A looming public health concern? Drug and Alcohol Dependence, 98(1–2), 1–12.
5. Morris, R. (n.d.). The effects of steroid use on depression among athletes [Unpublished manuscript]. Vital Quests, Associates for Health & Performance Improvement, Yonkers, NY. Archived at https://web.archive.org/web/20141026010602/http://vitalquests.org/publication2steroiduse.html
6. Perry, P. J., Andersen, K. H., & Yates, W. R. (1990). Illicit anabolic steroid use in athletes: A case series analysis. American Journal of Sports Medicine, 18(4), 422–428.
7. Pope, H. G. Jr., Gruber, A. J., Choi, P., Olivardia, R., & Phillips, K. A. (1997). Muscle dysmorphia: An underrecognized form of body dysmorphic disorder. Psychosomatics, 38(6), 548–557.
8. Pope, H. G. Jr., & Katz, D. L. (1988). Affective and psychotic symptoms associated with anabolic steroid use. American Journal of Psychiatry, 145(4), 487–490.
9. Pope, H. G. Jr., & Katz, D. L. (1994). Psychiatric and medical effects of anabolic-androgenic steroid use: A controlled study of 160 athletes. Archives of General Psychiatry, 51(5), 375–382.
10. Pope, H. G. Jr., Kouri, E. M., & Hudson, J. I. (2000). Effects of supraphysiologic doses of testosterone on mood and aggression in normal men: A randomized controlled trial. Archives of General Psychiatry, 57(2), 133–140.
11. Pope, H. G. Jr., Wood, R. I., Rogol, A., Nyberg, F., Bowers, L., & Bhasin, S. (2014). Adverse health consequences of performance-enhancing drugs: An Endocrine Society scientific statement. Endocrine Reviews, 35(3), 341–375.
12. Tricker, R., O'Neill, M. R., & Cook, D. (1989). The incidence of anabolic steroid use among competitive bodybuilders. Journal of Drug Education, 19(4), 313–325.