Relapse is one of the major challenges in the early stages of psychosis. For someone experiencing a First Episode of Psychosis (FEP), a relapse can involve a significant worsening of symptoms, increased clinical support or hospitalisation, and disruption to their recovery. This raises an important question: can we identify early signs that someone may be at greater risk of relapse?
Our recent study explored whether something as everyday sleep could provide part of the answer.
Sleep disturbance is very common in people experiencing First Episode Psychosis. Previous research has linked disrupted sleep with worsening psychosis symptoms, emotional difficulties and problems with everyday functioning. However, sleep is not simply a matter of sleeping “well” or “badly”. People may struggle to fall asleep, wake frequently during the night, sleep for too few hours, or experience substantial changes in how long they sleep from one night to the next.
We wanted to understand whether these different aspects of sleep could tell us something about a person's risk of relapse over the following year.
Following sleep in everyday life
The study involved 269 people experiencing First Episode Psychosis, recruited from ten NHS sites across the UK. During the first 30 days of a 12-month follow-up period, participants used a smartphone application to report on their sleep each day. Each assessment took approximately two to three minutes.
Participants were asked four simple questions: whether they had difficulty falling asleep, whether they woke frequently during the night, whether they felt they had obtained enough sleep, and how many hours they had slept.
This gave us a picture of people's sleep during the early stages of their involvement with the study. We could then examine whether these sleep patterns were related to whether participants subsequently experienced a relapse during the following 12 months.
What did we find?
The findings suggested that sleep contains useful information about subsequent relapse risk.
In particular, people who reported not getting enough sleep and having more nights of short sleep were more likely to be identified as being at greater risk of relapse. Greater irregularity in sleep duration was also associated with increased predicted risk. In contrast, longer average sleep duration was associated with lower predicted risk.
Importantly, it was not simply the amount of sleep that appeared to matter. The stability of sleep also seemed to be important. This suggests that disrupted or inconsistent sleep patterns may provide information about a person's clinical vulnerability.
The findings also suggested that sleep information collected over a relatively short period — just the first month — could provide information about what might happen over the following year. This is potentially important because sleep can be monitored relatively easily and repeatedly in everyday life.
Why might sleep be important?
One possibility is that sleep disturbance may be an early sign that someone is becoming clinically unwell.
Sleep and psychosis may influence one another in complex ways. Difficulties sleeping can be associated with increased stress, changes in mood and difficulties regulating emotions. Previous research has also suggested that losing sleep can contribute to increases in suspiciousness and hallucinations.
This means that changes in sleep could potentially provide an early warning sign that someone is becoming more vulnerable to worsening symptoms.
However, our study cannot tell us that poor sleep causes relapse. Sleep disturbance may instead be a marker of emerging clinical difficulties, or the relationship may work in both directions. Further research is needed to understand these processes.
What could this mean for clinical practice?
One of the most interesting aspects of these findings is that sleep is potentially modifiable.
If someone receiving care in an Early Intervention in Psychosis service is experiencing significant or changing sleep difficulties, this may be something worth exploring as part of their wider clinical assessment. Sleep could potentially become one part of a broader picture of how someone is doing, alongside changes in their mental health, stress levels, functioning and other early warning signs.
There may also be opportunities to intervene. Psychosocial interventions, such as Cognitive Behavioural Therapy for Psychosis (CBTp) can be used to address sleep difficulties, including strategies targeting sleep routines, unhelpful beliefs about sleep, insomnia, worry and hyperarousal.
However, our study did not test whether treating sleep reduces relapse, so this is an important question for future research.
What happens next?
There are several important steps before these findings could be applied in routine clinical practice.
Our findings need to be tested in new and independent groups of people experiencing First Episode Psychosis to establish whether the same patterns of findings are replicated in other samples. We also relied on people's own reports of their sleep, so future studies could combine this with information from wearable devices or other objective measures of sleep.
It would also be valuable to look at sleep alongside other aspects of people's everyday lives. Future research could combine sleep with information about activity, mood, cognition and other behavioural or physiological signals to explore whether these provide a more complete picture of emerging relapse risk.
We are also interested in understanding what happens psychologically when sleep becomes disrupted. For example, worry, rumination, intrusive thoughts and nightmares may all be relevant pathways linking sleep difficulties with psychosis experiences. Understanding these processes could help us develop more targeted psychological interventions.
The bigger picture
Ultimately, the goal is not simply to predict who might relapse. The more important question is whether identifying increased risk earlier can help us provide better and more personalised care.
Our findings suggest that everyday sleep patterns may provide useful information about relapse risk in First Episode Psychosis. Sleep is relatively easy to monitor, potentially modifiable, and closely connected with psychological wellbeing. This makes it an interesting target for future research into personalised relapse prevention.
One of the most interesting implications of this work is that something as ordinary as how we sleep may contain clues about how someone's mental health is changing. The next step is to find out how we can use those clues responsibly to support people experiencing early psychosis.