Psychotic symptoms, which encompass hallucinations and delusions, can be a feature of severe forms of depression. This psychotic depression can be found in both people with major depressive disorder (MDD, or ‘unipolar’ depression) and bipolar disorder (BD). Rough estimates suggest that it could affect up to as many as 1 in 100 people in their lifetime (Jääskeläinen et al., 2018), although these figures are based on community samples and a small number of studies. Despite its clinical importance, there are few studies that examine the long-term outcomes of psychotic depression, which makes it hard to determine the prognosis (i.e., what will happen) of people who experience it.
One way to get a more comprehensive view of prognosis in psychotic depression is by performing a systematic review, in which the findings of all previous studies on the topic are brought together. Previously, only a systematic review of prognosis in psychotic depression for patients with MDD had been carried out (Jääskeläinen et al., 2018), which found that psychotic depression was associated with greater depression severity and worse outcomes. Andreu and colleagues (2026) wanted to repeat this using up-to-date studies, as well as including studies on patients with BD, with the aim of providing “a more nuanced, prognosis- focused understanding of psychotic depression course and severity”.
Methods
Andreu et al. (2026) followed a set of instructions for carrying out a systematic review using a document called the ‘PRISMA statement’, which is considered the ‘gold-standard’ for performing a systematic review.
They searched three databases to identify all studies that included patients who had experienced psychotic depression as part of major depressive disorder (MDD) or bipolar disorder (BD), compared against patients with MDD or BD without psychotic symptoms. There also had to be some sort of measure of prognosis, such as suicidality, how well patients responded to treatment, or rates of relapse.
Three different researchers independently carried out the screening and extraction of information from the studies identified in the search, to ensure that no important details were missed in the final review.
The quality of the included studies was measured using the relevant methodological Joanna Briggs Institute critical appraisal tool. Around two-thirds of studies were judged to be at ‘low risk of bias’. However, this should not be confused with strong overall evidence: most studies were observational, and around a third were classified as providing relatively weak observational evidence.
Results
Forty-one studies were included in the final review, including data from 3,414 patients with psychotic depression and 11,763 with non-psychotic major depressive disorder (MDD) or bipolar disorder (BD). Most studies were observational (n = 24), with a minority focusing on the elderly (n = 6) or young people (n = 3).
Sociodemographic characteristics
Some of the included studies looked at certain sociodemographic factors such as marital status, education, and employment to see if they were associated with psychotic depression. However, no consistent results were found.
Clinical characteristics
Similar to sociodemographic factors, no consistent results were found regarding the relationship between the age of illness onset and psychotic depression. Findings regarding psychotic depression’s association with cognitive difficulties were also mixed.
Psychotic depression tended to be associated with more severe forms of depression when symptoms were measured using validated scales. Although the number of hospital admissions with psychotic depression varied across studies, psychotic depression tended to be associated with longer stays in hospital than non-psychotic depression. There also tended to be a higher rate of relapse, recurrence and rehospitalisation in psychotic depression patients. Correspondingly, psychotic depression patients tended to have lower rates of illness remission, and some studies found that they were less responsive to treatment.
Across studies, it was inconclusive whether patients with psychotic depression were more likely to ‘switch’ their diagnosis to BD at a later point (having originally been diagnosed with MDD). However some studies found that psychotic depression was associated with other psychiatric illnesses such as anxiety disorders, substance use disorders, and ‘personality disorders’.
Psychotic depression also tended to be associated with greater suicidality: patients tended to have more previous suicide attempts, higher levels of suicidal ideation (thinking about suicide), and a higher risk of suicide.
Functional outcomes
Some studies indicated that people with psychotic depression were more likely to be out of work, on benefits, or have difficulty carrying out work, or have impairments in educational settings. However, other studies indicated that there was no particular economic impairment in psychotic depression.
Treatment patterns
In terms of treatment, psychotic depression patients were more likely to receive intensive treatments, such as electroconvulsive therapy (ECT). Interestingly, some studies suggested that psychotic depression patients responded particularly well to ECT compared with non-psychotic depression patients, although the evidence was limited. Psychotic depression patients were also more likely to receive more than one pharmacological therapy at a time, in particular, combinations of antidepressants and antipsychotics.
Mortality
Finally, a small number of studies looked at mortality; results were mixed. Some indicated that the mortality rate was higher with psychotic depression compared to non-psychotic depression, others did not find such evidence.
Conclusions
The authors concluded that people with psychotic depression tend to have worse prognoses than those with non-psychotic depression. Across studies, individuals with psychotic features are more likely to need inpatient care, relapse more often, are more at risk of suicide, and have poorer day‑to‑day functioning. The authors stress the importance of identifying psychotic symptoms early, monitoring suicide risk and functioning closely, and intensifying treatment when needed. They also suggest that there needs to be more and larger studies into psychotic depression given some of the mixed findings in this review, leading them to reiterate that,
the available evidence supports the identification of general risk patterns and trends associated with psychotic features, rather than definitive or uniform prognostic effects.
Strengths and limitations
This was generally a well-carried out systematic review, with sound methodology. They ensured the robustness of their search and findings by:
- Registering the review before they carried it out; this ensures transparency and makes it harder, for instance, to be selective about which studies to include if they don’t fit a particular narrative.
- Using broad search terms to make sure that they didn’t miss any important papers in their search.
- Using three study databases to reduce the risk of missing important studies; they also checked the citation lists of included studies to widen their search.
- Having three separate researchers perform the screening and data extraction independently, with a fourth researcher resolving disagreements. This means that papers weren’t included/excluded on the whims of one person.
- Using validated critical appraisal tools to check the quality of the studies that they included in their review. The majority of studies were of good quality, which reduces our concerns about some sources of methodological bias.
There were some limitations, several of which were recognised by the authors themselves, and generally relate to the studies that were included rather than the review itself:
- Most included studies only looked at a small number of patients, which makes it hard to generalise their findings.
- Included studies didn’t always correct for other factors that may have been affecting the outcomes that were being explored (i.e., confounders). This weakens our confidence in the strength of the relationship between psychotic depression and those outcomes.
- Although the review included both major depressive disorder (MDD) and bipolar disorder (BD), most of the evidence came from people with MDD. Psychotic symptoms may have different prognostic implications in BD, so we should be cautious about assuming that the overall findings apply equally to both conditions.
- The review is billed as being an overview of prognosis in psychotic depression, i.e., what will happen in the future. Many of the included studies reported on snapshots of what patients were experiencing at the time (a ‘cross-sectional’ study); this certainly provides us with more information but may muddy the water in terms of predicting what lies in store for psychotic depression patients.
- The studies included in this review were written across a range of timepoints and in different parts of the world; as such, the findings of different studies may depend on different clinical contexts where, for instance, the threshold for hospitalisation or treatment pathways differ.
- Finally, the findings were synthesised narratively rather than in a meta-analysis. Because the studies were too different in their methods, populations and outcomes to statistically pool their results, we do not have an overall estimate of how strongly psychotic depression is associated with poorer outcomes, making the overall conclusions less precise.
Implications for practice
This is a comprehensive review on the prognosis of psychotic depression that highlights the potential severity of the condition across multiple important domains. Intriguingly, the findings of the review indicate that suicidal behaviour and thinking increases in psychotic depression, but that the conclusions regarding mortality in psychotic depression are much more uncertain. This might seem counter-intuitive (surely higher suicidality should translate to higher mortality?) but it could be the case that people with psychotic depression are more ‘obviously’ unwell and, as such, taken more seriously by clinicians. The review tells us that people with psychotic depression are hospitalised at greater rates than people with non-psychotic depression. Perhaps this protects them from other dangerous features of depression, such as self-neglect? However, this is just speculation; the authors note that these mixed findings are based on a small number of studies and, as such, further research into the link between psychotic depression and early death should be pursued.
The authors’ conclusions that emphasise the importance of identifying the condition early so that its damaging consequences can be limited is one that echoes what I have seen in clinical practice. I remember being a junior doctor on placement with a liaison psychiatry team (i.e., a psychiatry team based in a general hospital) and being asked to review an elderly woman with delirium – a condition commonly seen in hospital patients in which they become confused in response to physical illness. The team looking after her had tried various treatments to improve her confusion, but to no avail. I reviewed her and reported back to my consultant that part of her confusion was a persistent (mistaken) belief that she was homeless. My consultant recognised that this, in fact, was not simply confusion but a ‘delusion of poverty’; a fixedly held belief commonly seen in patients with psychotic depression, particularly in the elderly.
Often when we imagine psychosis, we picture a presentation that is florid, disturbing, and unignorable. This episode taught me that psychosis in depression, much like the broader illness, can be quiet and insidious.
With this in mind, the main take-away message of this review is that clinicians should be vigilant not to miss psychotic depression within MDD and BD. Simply put, people with psychotic depression appear to be at greater risk of poorer outcomes, although this was not consistent across every outcome or study. The higher rates of hospitalisation seen in psychotic depression may partly reflect the greater clinical complexity and risk experienced by this group, and, given that people with psychotic depression are more likely to become unwell again in the future, this vigilance should not end at recovery.
Statement of interests
Luke Baxter has no conflicts of interest to declare. AI was used to help with editing to improve flow and clarity.
Edited by
Dr Nina Higson-Sweeney.
Links
Primary paper
Helena Andreu, Vincenzo Oliva, Iñaki Ochandiano, Guadalupe Cabeza, Anna Giménez‐Palomo, Luis Olivier, Òscar de Juan, Lidia Ilzarbe, Eduard Vieta, & Isabella Pacchiarotti (2026). Psychotic Symptoms in Unipolar and Bipolar Depression and Their Impact on Disease Prognosis: A Systematic Review. Acta Psychiatrica Scandinavica, 154(1), 6-35. https://doi.org/10.1111/acps.70089
Other references
Jääskeläinen, E., Juola, T., Korpela, H., Lehtiniemi, H., Nietola, M., Korkeila, J., & Miettunen, J. (2018). Epidemiology of psychotic depression–systematic review and meta-analysis. Psychological Medicine, 48(6), 905-918. https://doi.org/10.1017/S0033291717002501