The physical effects of anabolic-androgenic steroids (AAS) are well documented, but their effects on mood, behaviour and sleep are just as real and often more disruptive to the people around a user than to the user themselves. This guide sets out what the human evidence actually shows, separates the well-supported findings from the caricatures, and is honest about where the data are weak. It is reference information, not medical advice.
The short version: AAS clearly affect mood in a meaningful minority of users, effects vary enormously between individuals, and the popular 'roid rage' story is both real and overblown depending on how you frame it. Sleep and dependence are underrated parts of the picture.
What the mood evidence shows
The most reliable evidence comes from a small number of controlled studies where testosterone or related androgens were given to healthy volunteers under blinded conditions, plus larger observational studies of users. Taken together, they support a few reasonably solid conclusions:
- A minority of users develop noticeable mood changes: irritability, mood swings, and in some cases hypomanic symptoms (elevated, expansive or agitated mood, reduced need for sleep, increased energy and impulsivity).
- The effect is dose-related. Controlled studies that pushed doses into the supraphysiologic range that mimics real cycles saw more mood disturbance than low, replacement-level doses, which are generally well tolerated.
- Response is highly individual. In blinded high-dose studies, most participants showed little psychological change, while a small subset developed marked hypomanic or aggressive symptoms. This variability is one of the most consistent findings: the same dose does very different things to different people.
Depression is the other side. Depressive symptoms are more commonly reported during withdrawal, after stopping, than on-cycle, which fits the hormonal crash that follows suppression (covered in the coming-off guide).
The honest state of the 'roid rage' literature
'Roid rage', the idea that steroids reliably turn users into violent aggressors, is a genuine phenomenon in some individuals but a poor description of the average effect. What the evidence supports:
- Controlled studies do show increases in self-rated and observer-rated aggression or irritability in some people at high doses, and case reports describe striking aggressive episodes.
- However, the effect is not universal or predictable. Many users show no meaningful change in aggression. Baseline personality, expectations, environment and pre-existing tendencies all matter, and some apparent 'roid rage' reflects people who were already prone to aggression, or who believed the drug would make them aggressive.
- The dramatic media framing of steroids as a direct, uniform cause of violence goes well beyond what the data support. The reality is a real but variable increase in irritability and aggression in a subset, not a switch that flips everyone.
The fair summary: steroids can increase irritability and aggression in susceptible individuals, especially at high doses, but the blanket 'roid rage' narrative overstates both the size and the consistency of the effect.
Dependence and behavioural effects
AAS dependence is real and under-recognised. A meaningful fraction of long-term users meet criteria for dependence: continued use despite harm, difficulty stopping, preoccupation with muscularity, and distress or dysphoria when off. Two drivers are worth separating:
- Body-image dependence: a compulsion to keep using driven by fear of losing size or by muscle dysmorphia.
- Neuroendocrine dependence: the low-testosterone withdrawal state after stopping is genuinely unpleasant, which pushes some men to resume rather than ride it out.
Dependence is more likely with longer use, higher doses, and where use is tied up with body image. It is one reason 'just stop' is harder than it sounds.
Sleep
Sleep is an under-discussed casualty. Several compounds and situations disrupt it:
- Hypomanic, high-energy states reduce the perceived need for sleep and can fragment it.
- Certain compounds, notably trenbolone and other potent androgens, are widely reported to cause insomnia, night sweats and vivid dreams; the mechanistic evidence is thinner than the anecdote, but the pattern is consistent enough to take seriously.
- Weight gain and higher training loads can worsen or unmask obstructive sleep apnoea, which itself degrades mood and daytime function.
Poor sleep then feeds back into irritability and low mood, so sleep disruption can amplify the other psychological effects rather than being separate from them.
Who is most at risk
Based on the controlled and observational data, higher-risk situations include: high supraphysiologic doses; a personal or family history of mood disorders, hypomania or aggression; use tied to body-image concerns or muscle dysmorphia; and the withdrawal period after stopping, when depressive symptoms peak. People with none of these often report little psychological change, which is exactly why the effects look so inconsistent across the literature.
When to get help
Seek help if: mood changes are affecting relationships, work or safety; you notice persistent aggression or hypomanic behaviour that others are commenting on; you feel unable to stop despite wanting to (a sign of dependence); or you develop significant depression, hopelessness or suicidal thoughts, particularly after coming off, which is when depressive symptoms are most common. A GP, psychiatrist, or a service experienced with performance-drug users can help; the withdrawal-related low mood in particular is treatable and time-limited for most people.
Bottom line
AAS affect mood, and in a susceptible minority they cause irritability, hypomania or increased aggression, more so at high doses; but the effect is individual and the blanket 'roid rage' story overstates it. Dependence is real and under-recognised, sleep disruption is common and amplifies everything else, and depression clusters around the withdrawal period rather than on-cycle. The evidence is strongest on the existence and variability of these effects and weakest on predicting who will be affected. Take mood changes seriously, watch the coming-off period especially, and get help if mood, behaviour or an inability to stop start causing harm.