Sleep Problems in Mental Illness Highly Pervasive

Kelli Whitlock Burton

October 15, 2021

An inpatient psychiatric diagnosis at some point over a lifetime is significantly associated with a range of sleep problems, results from the largest study of its kind show.

A prior diagnosis of major depression, schizophrenia, anxiety, or bipolar disorder was associated with a later bedtime, earlier waking time, and significantly poorer sleep quality that included frequent awakenings during the night and shorter sleep bouts.

Dr Michael Wainberg

"We were struck by the pervasiveness of sleep problems across all the diagnoses of mental illness and sleep parameters we looked at," study investigator Michael Wainberg, PhD, a postdoctoral fellow at the Krembil Centre for Neuroinformatics at the Center for Addiction and Mental Health  (CAMH), Toronto, Canada, told Medscape Medical News. "This suggests there may need to be even more of an emphasis on sleep in these patients than there already is."

The study, which includes data from nearly 90,000 adults in the United Kingdom, was published online October 12 in PLoS Medicine.

Trove of Data

Data for the analysis comes from the UK Biobank, a large-scale biomedical database launched in 2006 that has collected biological and medical data on more than 500,000 individuals who consented to provide blood, urine, and saliva samples and detailed lifestyle information that is matched to their medical records.

Between 2013 and 2015, more than 103,000 of these participants agreed to wear accelerometers on their wrists for 24 hours a day for 7 days, collecting a trove of data for researchers to mine.

Dr Shreejoy Tripathy

"This allows us to get at objectively derived sleep measures and to measure them in greater numbers of people who have experienced mental illness," said senior author Shreejoy Tripathy, PhD, assistant professor at the University of Toronto and independent scientist for CAMH. "You can study multiple disorders at once and the influence of other variables that might not be possible in the context of other studies."

The research is the first known large-scale transdiagnostic study of objectively measured sleep and mental health. Insomnia and other sleep disorders are common among people with mental illness, as shown in prior research, including at least one study that used the same dataset the team employed for this project.

The new findings add to that body of work, Wainberg said, and look beyond just how long a person sleeps to the quality of the sleep they get.

"We found that the metrics of sleep quality seem to be affected more than mere sleep duration," he said.

Unexpected Finding

After excluding participants with faulty accelerometers and those who didn't wear them for the entire 7-day study period, data from 89,205 participants (aged 43-79, 56% female, 97% self-reported White) was included. Lifetime inpatient psychiatric diagnoses were reported in 2.5% of the entire cohort.

Researchers looked at 10 sleep measures: bedtime, wake-up time, sleep duration, wake after sleep onset, sleep efficiency, number of awakenings, duration of longest sleep bout, number of naps, and variability in bedtime and sleep duration.

Although the effect sizes were small, having any psychiatric diagnosis was associated with significantly lower scores on every sleep measure except sleep duration.

Compared with those with no inpatient psychiatric diagnosis, those with any psychiatric diagnosis were significantly more likely to:

  • have a later bedtime (β = 0.07; 95% CI, 0.06 - 0.09)

  • have later wake-up time (β = 0.10; 95% CI, 0.09 - 0.11)

  • wake after sleep onset (β = 0.10; 95% CI, 0.09 - 0.12)

  • have poorer sleep efficiency (β = –0.12; 95% CI, −0.14 to −0.11)

  • have more awakenings (β = 0.10; 95% CI, 0.09 - 0.11)

  • have shorter duration of their longest sleep bout (β = –0.09; 95% CI, −0.11 to −0.08)

  • take more naps (β = 0.11; 95% CI, 0.09 - 0.12)

  • have greater variability in their bedtime (β = 0.08; 95% CI, 0.06 - 0.09)

  • have greater variability in their sleep duration (β = 0.10; 95% CI, 0.09 - 0.12)

The only significant differences in sleep duration were found in those with lifetime major depressive disorder, who slept significantly less (β = −0.02; P = .003), and in those with lifetime schizophrenia, who slept significantly longer (β = 0.02; P = .0008).

Researchers found similar results when they examined patient-reported sleep measures collected when participants enrolled in the biobank, long before they agreed to wear an accelerometer.

"Everyone with a lifetime mental illness diagnosis trended toward worse sleep quality, regardless of their diagnosis," Tripathy said. "We didn't expect to see that."

Limitations of the biobank data prohibited analysis by age and past or current use of psychiatric medications. In addition, investigators were unable to determine whether mental illness was active or controlled at the time of the study. Information on these, and other factors, is needed to truly begin to understand the real-world status of sleep patterns in people with mental illness, the researchers note.

However, the biobank data demonstrates how this type of information can be collected, helping Tripathy and others to design a new study that will launch next year with patients at CAMH. This effort is part of the BrainHealth Databank, a project that aims to develop a patient data bank similar to the one in the UK that was used for this study.

"We've shown that you can use wearable devices to measure correlates of sleep and derive insights about the objective measurements of sleep and associate them with mental illness diagnosis," Tripathy said.

The study received no outside funding. Wainberg and Tripathy report receiving funding from Kavli Foundation, Krembil Foundation, CAMH Discovery Fund, the McLaughlin Foundation, NSERC, and CIHR. Disclosures for other authors are fully listed in the original article.

PLoS Med. Published online October 12, 2021. Full text

For more Medscape Psychiatry news, join us on Twitter and Facebook


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.