Sleep Therapy Works Across Diagnoses, Even for Patients With Multiple Mental Illnesses

For people living with serious mental illness, sleep problems are rarely a side issue. They are often the thread that ties an entire clinical picture together, worsening mood, fueling psychotic symptoms, and eroding the ability to function day to day. A new study published in Community Mental Health Journal offers some of the strongest evidence yet that a single, sleep-focused treatment can help patients across the psychiatric spectrum, regardless of how many diagnoses they carry. The findings challenge a persistent worry in clinical practice: that patients burdened with multiple psychiatric disorders are simply too complex to benefit from a standardized psychological intervention.

The study, led by Tanya B. Horwitz and Allison G. Harvey of the University of California, Berkeley, is a secondary analysis of a randomized controlled trial of the Transdiagnostic Intervention for Sleep and Circadian Dysfunction, known as TSC. The researchers examined data from 489 patients treated in Community Mental Health Centers across ten California counties. All participants had serious mental illness, defined as at least one DSM-5 disorder causing substantial interference with major life activities for at least a year, along with significant sleep or circadian problems. Crucially, the sample was clinically diverse: patients reported an average of 2.29 psychiatric disorders, and 76 percent reported at least two.

TSC itself is a modular treatment designed to target the sleep and circadian mechanisms that cut across diagnostic boundaries. Rather than treating insomnia as a symptom of depression or mania, TSC addresses the underlying sleep health directly. The Standard version consists of eight 50-minute weekly sessions, including four cross-cutting modules such as motivation enhancement, four core modules covering the basic building blocks of sleep health like regular sleep-wake timing, and seven optional modules tailored to individual presentations, such as reducing sleep-related worry. An Adapted version, developed specifically for the under-resourced community mental health context, condenses this into four 20-minute weekly sessions with a single optional module.

The trial assigned patients either to receive TSC immediately alongside their usual care, or to a comparison group receiving usual care first, followed by delayed treatment. Sites were cluster-randomized by county to deliver either the Standard or Adapted version. The researchers then asked a deceptively simple question: did the number of psychiatric disorders a patient reported predict how impaired they were at baseline, and did it predict how much they improved?

The first answer was a clear yes. Patients with more diagnoses arrived at treatment in worse shape across nearly every measure. Each additional disorder was associated with significantly higher scores on the DSM-5 Level 1 Cross-Cutting Symptom Measure, a 23-item questionnaire spanning 13 psychiatric symptom domains. More diagnoses also predicted significantly greater sleep-related daytime impairment, worse overall sleep health on a six-dimension composite covering regularity, timing, efficiency, duration, satisfaction, and alertness, and greater functional impairment on the Sheehan Disability Scale. The relationship with sleep disturbance itself was positive but only marginally significant after conservative outlier removal. These results echo the National Comorbidity Survey Replication, which found that nearly half of respondents with three or more psychiatric disorders qualified as serious cases, compared with fewer than ten percent of those with a single disorder.

The second answer, however, was a resounding no, and this is where the study becomes genuinely striking. The number of psychiatric disorders did not predict treatment outcome at all. Patients with many comorbidities improved just as much as those with fewer. Nor did diagnostic count determine whether patients fared better with the Standard or the Adapted version of TSC. All interaction tests examining whether comorbidity moderated treatment benefit came back non-significant. In other words, carrying multiple diagnoses, which might seem like an obvious reason to expect poorer response, made no measurable difference to how much patients gained from sleep-focused treatment.

The researchers then went a step further, analyzing the five most common broad diagnostic groups independently: unipolar mood features, bipolar mood features, anxiety disorders, psychosis, and trauma and stressor-related disorders. The results were remarkably consistent. Every one of the five groups showed significantly more improvement in the immediate TSC condition than in the usual care comparison group on both sleep disturbance and sleep-related impairment, even after applying statistical corrections for multiple testing. Patients with unipolar mood features and anxiety disorders, the two largest groups, improved significantly across all five outcomes, including psychiatric symptoms, overall sleep health, and functional impairment. The psychosis group showed significant gains on sleep measures, sleep health, and daily functioning, with psychiatric symptom improvement reaching the nominal threshold. The trauma and stressor-related group improved significantly on all outcomes except the cross-cutting symptom measure.

These findings matter because sleep is increasingly understood as a mechanistic driver of psychiatric illness rather than a mere byproduct. Sleep and circadian disruption has been implicated in the worsening of delusions and hallucinations, is a common prodrome of mania in bipolar disorder, and predicts the development of post-traumatic stress disorder after traumatic events. Poor sleep and psychiatric symptoms can feed each other in destructive loops: a patient may use substances to fall asleep and stimulants to fight daytime fatigue, deepening both the sleep problem and the risk of addiction. By targeting sleep directly, TSC appears to interrupt these loops at a shared point of vulnerability, which may explain why a single treatment produces benefits across such clinically different populations.

The study also carries a quieter but important message about measurement. Because provider-reported diagnostic data were missing for many patients, the researchers relied primarily on patients’ own reports of their diagnoses. Far from being a fatal weakness, this choice was supported by prior evidence: self-reported diagnostic accuracy has been shown to reach 70 percent or higher for most psychiatric disorders, and polygenic risk scores for schizophrenia have been found to be essentially equivalent whether diagnoses come from self-report, research interviews, or medical records. The finding that self-reported diagnostic counts meaningfully predicted baseline impairment suggests that patients’ knowledge of their own psychiatric histories is a usable and resource-efficient clinical signal, one that aligns with person-centered approaches to care.

Limitations remain. The analyses of treatment moderation may have been underpowered, since detecting three-way statistical interactions requires substantially larger samples. Post-treatment assessments had considerable missing data, session counts varied between patients, and the clinical diversity that makes transdiagnostic research possible also makes interpretation harder. Still, the bottom line is hopeful and practical: patients with more psychiatric disorders, despite starting treatment with more severe symptoms and greater impairment, can be strong candidates for sleep-focused transdiagnostic care. A high diagnostic count is not an insurmountable barrier to recovery. For community mental health centers serving low-income populations with limited resources, that is a message worth acting on, because it means the patients who are often considered the most complicated may be exactly the ones this treatment can reach.

Subject of Research: Transdiagnostic sleep and circadian treatment for serious mental illness and psychiatric comorbidity in community mental health settings

Article Title: The Transdiagnostic Intervention for Sleep and Circadian Dysfunction (TSC) in Community Mental Health: Evaluating Self-Reported Psychiatric Disorders as a Predictor of Symptoms and Treatment Outcome

Article References: B. Horwitz, T., D. Sarfan, L., E. Milner, A., Varghese, J., A. Callaway, C., & G. Harvey, A. (2026). The Transdiagnostic Intervention for Sleep and Circadian Dysfunction (TSC) in Community Mental Health: Evaluating Self-Reported Psychiatric Disorders as a Predictor of Symptoms and Treatment Outcome. Community Mental Health Journal. https://doi.org/10.1007/s10597-026-01719-1

Image Credits: AI Generated

DOI: 10.1007/s10597-026-01719-1

Keywords: sleep disorders, circadian rhythms, serious mental illness, psychiatric comorbidity, transdiagnostic treatment, community mental health, randomized controlled trial, insomnia, self-reported diagnoses, TSC intervention, treatment outcomes, psychosis