Incidence and mechanisms of action

Approximately 5% of patients treated with corticosteroids experience severe psychiatric reactions. A literature review found that the median time to symptom onset is 11.5 days; 39% develop symptoms in the first week and 62% within 2 weeks. Symptoms may also start on discontinuation of corticosteroids (Lu Reference Lu, Ann and McCarron2021). Corticosteroids are clinically indicated for numerous health conditions, but it is also important for clinicians to be aware of their potential misuse, for example to enhance athletic performance in certain settings. Individuals misusing corticosteroids may also present to services with neuropsychiatric adverse reactions.

Higher corticosteroid doses are the most important risk factor for psychiatric adverse effects. However, dosage has not been linked to onset, severity, reaction type or duration (Huynh Reference Huynh and Reinert2020).

The pathophysiology is unknown; theories link the phenomenon to reduced levels of corticotropin, noradrenaline and beta-endorphin immunoreactivity, and possible effects on areas such as the hippocampus and amygdala (Lu Reference Lu, Ann and McCarron2021).

Diagnosis and management

Diagnosis is through clinical assessment, although there are no specific confirmatory investigations; thorough information gathering, including medication history, is essential. Liaison psychiatry teams play an integral role in acute hospitals. The best treatment setting (acute hospital, psychiatric hospital or community) is determined by the severity/complexity of the patient’s physical and mental health presentation. Early identification is key, and management should take an individualised and integrated approach.

• Close working between relevant medical specialties and psychiatry is necessary to treat the primary presenting complaint alongside any new psychiatric symptoms.

• Clear, concise and timely communication, including cross-specialty multidisciplinary team (MDT) meetings, is vital to reduce misinformation resulting from multiple team involvement.

• Regular cross-specialty MDT meetings should be organised at a frequency determined by symptom severity/complexity; initially, they may be needed weekly in more complex/unwell patients.

• Attendees should include the medical teams treating the physical illness and the most appropriate psychiatric team(s), such as liaison psychiatry, in-patient psychiatry if psychiatric admission is needed or crisis/community psychiatry if the patient is at home.

• Patients should be included in shared management decisions where possible, and the most appropriate legal framework must be considered if mental capacity is impaired.

Management is based on clinical judgement alongside limited evidence. No formal treatment guidelines or randomised control trials exist for corticosteroid-induced psychosis. Available evidence primarily includes retrospective cohort studies, case–control studies and case series, with limitations including possible biases, limited generalisability and difficulties establishing causal relationships. Of the references cited below, Gagliardi et al (Reference Gagliardi, Muzyk and Holt2010) and Goel (Reference Goel, Pickard and Hindmarsh2025) are from non-peer-reviewed sources.

There is no definitive medical management other than reducing corticosteroids to the lowest possible dose and stopping them if possible (Goel Reference Goel, Pickard and Hindmarsh2025). Adjunctive psychotropic management includes the following.

• Antipsychotics – Both first- and second-generation antipsychotics can be used, including haloperidol, olanzapine and risperidone. It is beneficial to start antipsychotics at low doses and titrate to the lowest therapeutic dose. Adding a second antipsychotic is not recommended owing to a lack of robust evidence supporting the efficacy of combined antipsychotics, with considerable evidence of potential harm from side-effects.

• Benzodiazepines – Short-term diazepam or lorazepam for behavioural disturbance and relief of distressing symptoms is reasonable, depending on symptom severity and the impact of the patient’s behaviour. Such medicines should be employed judiciously at the lowest effective dose, for the shortest possible time, to mitigate risk of harm.

• Mood stabilisers – A mood stabiliser such as valproate, carbamazepine or lithium could be considered, especially when the clinical picture is dominated by manic/hypomanic symptoms. Psychotropic polypharmacy is generally not advised, however, and there is a potential for drug–drug interactions with enzyme inhibitors such as valproate and carbamazepine. The use of these medications can also be complicated in medically ill patients. Many patients prescribed corticosteroids have comorbidities that affect renal and/or hepatic function. Therefore, lithium and valproate may be difficult to use safely, and close renal and hepatic monitoring would be required respectively (Gagliardi Reference Gagliardi, Muzyk and Holt2010).

• Antidepressants – Antidepressants may be considered when significant depressive symptoms are reported. It is important to screen for a history of mania prior to initiation. Caution is required as many antidepressants, including selective serotonin and selective noradrenaline reuptake inhibitors (SSRIs and SNRIs) are metabolised through the cytochrome P450 iso-enzyme family, leading to potential drug–drug interactions. They may cause adverse effects, including the syndrome of inappropriate antidiuretic hormone secretion (SIADH), increased bleeding risk and QTc prolongation, especially in elderly patients (Gagliardi Reference Gagliardi, Muzyk and Holt2010). Tricyclic agents can exacerbate agitation and anxiety in individuals with corticosteroid-induced psychosis and should therefore be avoided (Lu Reference Lu, Ann and McCarron2021). Consider electroconvulsive therapy (ECT) for patients with severe depression.

Clinicians must be cautious when prescribing corticosteroids, so as to reduce the risk of preventable psychiatric adverse drug reactions. It may take days to weeks before symptoms improve, but most patients experiencing corticosteroid induced psychotic symptoms generally recover from their acute presentation within 2 weeks of starting appropriate management (Huynh Reference Huynh and Reinert2020).