Sleep Disturbances as Psychiatric Risk Markers: Insomnia, Hypersomnia, and Their Prognostic Significance in Mental Health

Sleep disturbances occupy a singular position in psychiatric nosology. Historically classified as mere symptoms—epiphenomena of underlying mood, anxiety, or psychotic disorders—insomnia and hypersomnia are increasingly recognized as independent risk markers, prodromal signals, and perpetuating mechanisms for a broad range of psychiatric conditions. This reconceptualization has been driven by longitudinal epidemiological studies, neurobiological research, and clinical trial data demonstrating that disordered sleep is not merely a downstream consequence of mental illness but frequently precedes, amplifies, and sustains it.

The DSM-5-TR reflects this paradigm shift by abandoning the older distinction between "primary" and "secondary" insomnia, replacing it with Insomnia Disorder as a standalone diagnosis that can be comorbid with—but not reducible to—other psychiatric conditions. The ICD-11 mirrors this framework with its classification of chronic insomnia (7A00) and hypersomnia disorders (7A20). This diagnostic evolution carries direct clinical implications: sleep disturbances warrant independent assessment and treatment even when they co-occur with depression, anxiety, PTSD, or psychosis.

The epidemiological scale is substantial. Approximately 10–15% of the general adult population meets criteria for chronic insomnia disorder, while an additional 25–35% reports occasional insomnia symptoms. Hypersomnia, though less prevalent at approximately 4–6% of the general population, carries its own distinct prognostic profile—particularly in the context of mood disorders, where it signals treatment complexity and recurrence risk. Among psychiatric outpatients, rates of clinically significant sleep disturbance exceed 50–80% depending on the disorder and assessment method, dwarfing general population estimates.

This article examines the neurobiology, diagnostic nuances, prognostic significance, and evidence-based treatment of insomnia and hypersomnia as psychiatric risk markers, synthesizing data from landmark studies and meta-analyses to provide a clinically actionable framework.

The Hyperarousal Model of Insomnia

The dominant neurobiological framework for chronic insomnia is the hyperarousal hypothesis, which posits that insomnia reflects a state of persistent physiological, cognitive, and cortical hyperactivation that extends across the full 24-hour period—not merely at night. Evidence for this model comes from multiple converging lines of research:

• HPA axis dysregulation: Individuals with chronic insomnia show elevated 24-hour cortisol levels, particularly in the evening and early nocturnal period when cortisol should reach its nadir. Meta-analytic data demonstrate a small but reliable elevation in mean 24-hour cortisol (Cohen's d ≈ 0.2–0.4) in insomnia versus good sleepers. This mirrors the HPA axis dysfunction seen in major depressive disorder (MDD), suggesting a shared neuroendocrine vulnerability.
• GABAergic deficits: Proton magnetic resonance spectroscopy (¹H-MRS) studies have demonstrated reduced gamma-aminobutyric acid (GABA) levels in the occipital cortex and anterior cingulate cortex of individuals with primary insomnia. Since GABA is the brain's principal inhibitory neurotransmitter and plays a critical role in sleep initiation and maintenance via GABA_A receptors in the ventrolateral preoptic (VLPO) nucleus of the hypothalamus, these deficits directly implicate inhibitory failure in insomnia pathophysiology.
• Orexin/hypocretin system: The orexin (hypocretin) system, originating in the lateral hypothalamus, serves as a wake-promoting stabilizer. Elevated orexin signaling contributes to the "flip-flop switch" becoming biased toward wakefulness—a mechanism exploited therapeutically by dual orexin receptor antagonists (DORAs) such as suvorexant and lemborexant.
• Cortical hyperarousal: EEG studies consistently show elevated beta-frequency (16–32 Hz) power during NREM sleep in insomnia patients, indicating inadequate cortical deactivation. Neuroimaging during the sleep-wake transition reveals persistent metabolic activity in the ascending reticular activating system (ARAS), thalamus, and prefrontal cortex—regions that normally show marked deactivation during sleep onset.

Hypersomnia: Distinct Neurochemical Signatures

Hypersomnia associated with psychiatric conditions—particularly the atypical depression subtype—involves different neurobiological substrates. Central hypersomnia (idiopathic hypersomnia, narcolepsy type 2) involves dysfunction in wake-promoting systems, but psychiatric hypersomnia may reflect:

• Reduced histaminergic tone: The tuberomammillary nucleus (TMN), a key wake-promoting center that uses histamine as its primary neurotransmitter, shows diminished output in states of hypersomnia. This underpins the wake-promoting action of histamine H₃ receptor inverse agonists such as pitolisant.
• Dopaminergic hypofunction: The ventral tegmental area (VTA) and substantia nigra pars compacta contribute to arousal and motivated wakefulness. Hypersomnia in depression correlates with blunted mesolimbic dopamine signaling, and stimulant medications (modafinil, methylphenidate) act partly through dopaminergic mechanisms.
• Endogenous somnogenic substances: Research by Rye and colleagues has identified a small-molecular-weight substance in the cerebrospinal fluid of idiopathic hypersomnia patients that acts as a positive allosteric modulator of GABA_A receptors, effectively enhancing GABA-mediated inhibition. This finding, while not yet fully replicated across all studies, suggests that some forms of hypersomnia involve excessive inhibitory signaling—the neurochemical mirror image of the GABAergic deficit in insomnia.

Shared Genetic Architecture

Genome-wide association studies (GWAS) have identified shared genetic loci between insomnia and psychiatric disorders. The landmark UK Biobank GWAS (Jansen et al., 2019), studying over 1.3 million participants, identified 202 risk loci for insomnia complaints, with significant genetic correlations (r_g) between insomnia and MDD (r_g = 0.44), anxiety disorders (r_g = 0.56), and neuroticism (r_g = 0.53). Mendelian randomization analyses from this and subsequent studies suggest that insomnia has a causal influence on depression risk, not merely a shared liability—a finding with profound implications for prevention.

Major Depressive Disorder

Insomnia is the single most common residual symptom following otherwise successful depression treatment and one of the strongest prospective predictors of depressive onset. The landmark meta-analysis by Baglioni et al. (2011), synthesizing 21 longitudinal studies with over 30,000 participants, found that non-depressed individuals with insomnia had a twofold risk (pooled OR = 2.10, 95% CI 1.86–2.38) of developing MDD over follow-up periods ranging from 1 to 34 years. This effect was robust across age groups, assessment methods, and study designs.

Critically, the STAR*D trial revealed that insomnia was among the most treatment-resistant symptoms across all four treatment steps. Even among remitters (defined by HAM-D ≤ 7), approximately 25–30% continued to report clinically significant insomnia symptoms, and these residual symptoms predicted relapse over subsequent follow-up. This has led to the clinical recommendation that insomnia should be treated as a co-target in depression management, not merely an expected correlate that resolves with antidepressant response.

Bipolar Disorder

Sleep disturbance in bipolar disorder has a bidirectional and phase-specific character. Insomnia—particularly a reduced need for sleep—is a hallmark of manic episodes and often the earliest prodromal symptom preceding full mania. The Bipolar Collaborative Network data show that sleep disturbance precedes manic episodes in approximately 77% of cases and depressive episodes in approximately 56% of cases, typically by 1–3 days. Conversely, hypersomnia is a prominent feature of bipolar depression, present in 38–78% of bipolar depressive episodes depending on the study, significantly higher than the 15–25% rate in unipolar depression. This differential prevalence of hypersomnia has clinical utility in differentiating bipolar from unipolar depression at initial presentation.

Post-Traumatic Stress Disorder (PTSD)

Insomnia and trauma-related nightmares are not merely associated features of PTSD; they appear to be mechanistically central to its development and maintenance. The prospective study by Gehrman et al. (2013) among military personnel demonstrated that pre-deployment insomnia symptoms significantly predicted post-deployment PTSD symptom severity, even after controlling for combat exposure, prior trauma, and baseline psychological distress. The current consensus, codified in the DSM-5-TR, is that sleep disruption impairs the consolidation and extinction of fear memories—processes dependent on REM sleep—thereby preventing the natural resolution of trauma responses.

Psychotic Disorders

Insomnia frequently precedes first-episode psychosis and psychotic relapse. In a meta-analysis of the ultra-high-risk (UHR) population, sleep disturbance was present in approximately 80% of individuals who subsequently converted to psychosis. In established schizophrenia, circadian rhythm disruption is near-universal, driven partly by dopamine-melatonin interactions and social zeitgeber disruption. The clinical implication is that acute-onset insomnia in a patient with known psychotic disorder warrants urgent reassessment, as it may herald relapse days to weeks before positive symptoms re-emerge.

Suicidality

Perhaps the most clinically urgent function of insomnia as a risk marker is its independent association with suicidal ideation, suicide attempts, and completed suicide. The meta-analysis by Pigeon et al. (2012), examining 39 studies, found that sleep disturbance was associated with a significantly elevated risk of suicidal behavior (OR ≈ 1.95 for insomnia and suicidal ideation; OR ≈ 2.70 for nightmares and suicidality), even after controlling for depression severity. This finding is clinically actionable because sleep disturbance is a modifiable risk factor—unlike many other suicide risk markers such as prior attempts or family history.

Hypersomnia is clinically underrecognized relative to insomnia, partly because patients and clinicians often normalize excessive sleepiness and partly because its differential diagnosis is complex. The DSM-5-TR defines Hypersomnolence Disorder as recurrent periods of sleep or irresistible need to sleep within the same day, occurring at least three times per week for at least three months, causing clinically significant distress or impairment. However, hypersomnia as a psychiatric symptom most commonly presents within the context of mood disorders, substance use, or medication effects rather than as a standalone diagnosis.

Hypersomnia in Mood Disorders: A Prognostic Signal

In unipolar depression, the presence of hypersomnia (typically defined as sleeping ≥9–10 hours or experiencing marked daytime sleepiness) is associated with:

• Earlier age of onset of depressive illness
• Greater chronicity: Depressive episodes with hypersomnia tend to be longer and more recurrent
• Poorer antidepressant response: The atypical depression subtype, characterized by hypersomnia, hyperphagia, leaden paralysis, and rejection sensitivity, shows differential medication response—historically responding better to MAOIs (phenelzine demonstrated superiority over imipramine in the Columbia Atypical Depression Study) and, in more recent practice, showing variable response to SSRIs
• Higher bipolar conversion rates: Among initially unipolar depressed patients, those with prominent hypersomnia have a higher rate of subsequent bipolar diagnosis (estimated at 20–30% over 10-year follow-up in some cohorts), making hypersomnia a valuable soft sign for bipolar diathesis

Differential Diagnosis Pitfalls

Clinicians evaluating psychiatric hypersomnia must systematically exclude:

• Obstructive sleep apnea (OSA): Present in approximately 2–4% of the general population but 17–52% of psychiatric populations depending on the sample, often undiagnosed. OSA contributes to both excessive sleepiness and treatment resistance in depression. Every patient with hypersomnia complaints should be screened with the STOP-BANG questionnaire or equivalent.
• Narcolepsy type 2: Characterized by chronic excessive daytime sleepiness without cataplexy, often misdiagnosed as depression or chronic fatigue for 10–15 years before correct identification. Multiple Sleep Latency Test (MSLT) showing mean sleep latency <8 minutes with ≥2 sleep-onset REM periods (SOREMPs) is the diagnostic standard.
• Idiopathic hypersomnia: Distinguished from narcolepsy by long, unrefreshing naps and sleep inertia (severe difficulty waking), typically without SOREMPs on MSLT. Recently approved treatment with low-sodium oxybate (Xywav) demonstrated significant reduction in Epworth Sleepiness Scale scores in the pivotal trial.
• Medication-induced hypersomnia: Sedating antidepressants (mirtazapine, trazodone, tricyclics), antipsychotics (quetiapine, olanzapine), benzodiazepines, and antihistamines are frequent culprits.
• Circadian rhythm disorders: Delayed sleep-wake phase disorder, common in adolescents and young adults, produces a clinical picture of apparent hypersomnia in the morning and insomnia at night that is often misdiagnosed as depression.

Accurate characterization of sleep disturbance in psychiatric populations requires a multimodal assessment approach. Reliance on a single screening question ("Are you sleeping okay?") is insufficient to capture the heterogeneity and clinical significance of sleep complaints.

Validated Instruments

• Insomnia Severity Index (ISI): A 7-item self-report measure with strong psychometric properties. Clinical cutoffs: 0–7 = no clinically significant insomnia; 8–14 = subthreshold; 15–21 = moderate clinical insomnia; 22–28 = severe clinical insomnia. The ISI has sensitivity of approximately 86% and specificity of approximately 88% for detecting insomnia disorder when using a cutoff of 10.
• Pittsburgh Sleep Quality Index (PSQI): A 19-item measure assessing seven components of sleep quality over the past month. A global score >5 identifies poor sleepers with a diagnostic sensitivity of 89.6% and specificity of 86.5% in its original validation by Buysse et al. (1989). Though widely used, it captures sleep quality broadly rather than insomnia severity specifically.
• Epworth Sleepiness Scale (ESS): The standard measure for daytime sleepiness, with scores >10 indicating excessive sleepiness and scores >15 indicating severe sleepiness. Essential for differentiating hypersomnia from fatigue (which does not improve with sleep).
• Sleep diaries: Two-week prospective sleep diaries remain the clinical standard for characterizing sleep-wake patterns, capturing night-to-night variability that single-point questionnaires miss. Key variables include sleep onset latency (SOL), wake after sleep onset (WASO), total sleep time (TST), and sleep efficiency (SE = TST/time in bed × 100; values <85% are generally considered clinically significant).

Objective Measures

Polysomnography (PSG) is not routinely indicated for insomnia assessment per AASM guidelines but is essential when sleep-disordered breathing, narcolepsy, or periodic limb movement disorder is suspected. PSG findings in psychiatric insomnia often include prolonged SOL, increased WASO, reduced sleep efficiency, and alterations in sleep architecture—particularly shortened REM latency (time from sleep onset to first REM period), which has been one of the most replicated biological markers in depression research, present in approximately 40–60% of MDD patients.

Actigraphy, using wrist-worn accelerometers, provides 1–2 weeks of objective sleep-wake data in the natural environment and is increasingly recommended for circadian rhythm disorders and for validating self-reported sleep duration. Actigraphy frequently reveals discrepancies between subjective and objective sleep in both insomnia (where patients often underestimate sleep obtained—so-called "sleep state misperception" or "paradoxical insomnia") and hypersomnia.

CBT for Insomnia (CBT-I): First-Line Treatment

Cognitive Behavioral Therapy for Insomnia (CBT-I) is the consensus first-line treatment for chronic insomnia per the American Academy of Sleep Medicine (AASM), the American College of Physicians (ACP), and the European Sleep Research Society (ESRS). It is a structured, typically 4–8 session intervention comprising five core components:

• Sleep restriction therapy: Reducing time in bed to match actual sleep time, thereby increasing homeostatic sleep drive and consolidating sleep
• Stimulus control: Re-associating the bed and bedroom with sleep rather than wakefulness
• Cognitive restructuring: Addressing catastrophic beliefs about sleep loss and maladaptive sleep effort
• Sleep hygiene education: Optimizing behavioral and environmental factors (though sleep hygiene alone is ineffective as a standalone intervention)
• Relaxation training: Progressive muscle relaxation, diaphragmatic breathing, or mindfulness techniques

Efficacy data: Meta-analyses consistently show large effect sizes for CBT-I on insomnia severity (Hedges' g = 0.98–1.09 vs. waitlist controls on ISI scores). The meta-analysis by Trauer et al. (2015) in the Annals of Internal Medicine found that CBT-I reduced sleep onset latency by approximately 19 minutes, reduced WASO by approximately 26 minutes, and improved sleep efficiency by approximately 10 percentage points. Critically, these improvements are durable—maintained or even enhanced at 6–12 month follow-up, unlike pharmacotherapy effects which typically dissipate upon discontinuation.

In head-to-head comparisons, the Mitchell et al. (2012) study comparing CBT-I versus zolpidem versus their combination found that CBT-I alone and the combination produced superior long-term outcomes compared to zolpidem alone, with CBT-I showing continued improvement after the active treatment period while medication-only participants reverted toward baseline.

CBT-I in Comorbid Psychiatric Conditions

The pivotal Lancet Psychiatry trial by Freeman et al. (2017)—the largest randomized controlled trial of a psychological intervention for insomnia (n = 3,755)—demonstrated that digital CBT-I (delivered via the Sleepio platform) not only improved insomnia (Cohen's d = 1.11) but also produced significant reductions in paranoia (d = 0.19) and hallucinatory experiences (d = 0.24) in a university student population. Though effect sizes on psychiatric outcomes were modest, they provided experimental evidence that insomnia treatment can produce downstream benefits on psychotic symptoms.

In comorbid insomnia and depression, the Manber et al. (2008) trial demonstrated that adding CBT-I to antidepressant treatment doubled the depression remission rate (HAM-D ≤ 7) from 33% to 62% compared to antidepressant plus control sleep intervention—one of the most striking demonstrations that treating insomnia independently improves mood outcomes.

Pharmacotherapy for Insomnia

When pharmacotherapy is indicated—typically for short-term use, when CBT-I is unavailable, or as adjunctive treatment during CBT-I initiation—several classes are available with varying evidence bases:

• Benzodiazepine receptor agonists (BzRAs): Zolpidem, eszopiclone, and zaleplon. Meta-analytic data show they reduce SOL by approximately 12–20 minutes and increase TST by 20–30 minutes versus placebo—statistically significant but clinically modest. NNT ≈ 6–13 for a clinically meaningful response, with a NNH ≈ 6–8 for side effects. Long-term use is complicated by tolerance, dependence, rebound insomnia, and associations with falls and cognitive impairment in older adults.
• Dual orexin receptor antagonists (DORAs): Suvorexant (10–20 mg) and lemborexant (5–10 mg) block orexin-1 and orexin-2 receptors, reducing wake drive without the GABAergic side-effect profile. Lemborexant demonstrated superiority over placebo and non-inferiority to zolpidem ER on WASO in the SUNRISE-1 and SUNRISE-2 trials, with maintained efficacy over 12 months without evidence of tolerance. DORAs may be particularly suitable for psychiatric populations given their favorable cognitive safety profile.
• Melatonin receptor agonists: Ramelteon (MT1/MT2 agonist) modestly reduces SOL (by ~7–16 minutes) with minimal abuse potential. Its effect size is smaller than BzRAs, but it carries no dependence risk.
• Sedating antidepressants: Trazodone (25–100 mg), doxepin (3–6 mg at the FDA-approved low dose), and mirtazapine are widely prescribed for insomnia, though evidence quality varies. Low-dose doxepin has the strongest evidence base in this category, with FDA approval for sleep maintenance insomnia based on rigorous RCT data showing improved WASO without morning sedation at low doses. Trazodone is the most commonly prescribed medication for insomnia in the United States despite a thin evidence base—only a handful of small RCTs, most of short duration.
• Suvorexant: In a secondary analysis of its pivotal trials, suvorexant showed particular benefit for insomnia comorbid with MDD, improving both sleep and daytime functioning.

Pharmacotherapy for Psychiatric Hypersomnia

Treatment options for hypersomnia in psychiatric contexts are less well-studied. Modafinil (100–400 mg) and armodafinil (150–250 mg) are the most commonly used wake-promoting agents, acting primarily through dopamine transporter (DAT) blockade. In adjunctive use for residual hypersomnia in treated depression, modafinil has shown modest efficacy (NNT ≈ 7–10 in available trials) with generally favorable tolerability, though it carries interaction risks with hormonal contraceptives and has limited long-term data in psychiatric populations. Solriamfetol, a dopamine and norepinephrine reuptake inhibitor approved for OSA and narcolepsy-related sleepiness, represents a newer option but lacks psychiatric-specific trial data.

Identifying which patients with sleep disturbance will respond to treatment—and which will develop persistent, treatment-resistant symptoms—is a clinical priority with significant research behind it.

Favorable Prognostic Factors for CBT-I Response

• Higher baseline insomnia severity: Paradoxically, patients with more severe insomnia often show larger absolute improvements with CBT-I, likely because sleep restriction therapy generates greater homeostatic sleep drive
• Treatment adherence: Adherence to sleep restriction and stimulus control prescriptions is the strongest predictor of CBT-I response. Patients who comply with time-in-bed restrictions show remission rates of 60–70% versus 25–30% for non-adherent patients
• Absence of hypnotic dependence: Patients on long-term benzodiazepines or Z-drugs can still benefit from CBT-I but typically require slower titration and longer treatment courses
• Intact circadian rhythmicity: Patients with preserved circadian amplitude (measurable via actigraphy or core body temperature) respond better than those with flattened circadian profiles

Poor Prognostic Factors

• Comorbid untreated OSA: Insomnia and OSA co-occur in 30–50% of clinical insomnia populations (the "COMISA" phenotype—Comorbid Insomnia and Sleep Apnea). COMISA patients have worse outcomes for both insomnia treatment and CPAP adherence compared to patients with either condition alone. The ETHOS trial has demonstrated that sequential treatment (CBT-I first, then CPAP) may improve outcomes for both conditions.
• Short sleep duration insomnia: Insomnia with objectively short sleep (PSG-confirmed TST <6 hours) is associated with greater physiological hyperarousal, higher cortisol, increased cardiometabolic risk, and poorer CBT-I response compared to insomnia with normal sleep duration. This phenotypic distinction, proposed by Vgontzas and Fernandez-Mendoza, may represent a more biologically severe insomnia subtype.
• Psychiatric comorbidity severity: While CBT-I is effective in comorbid psychiatric conditions, severe, untreated psychiatric illness (particularly active psychosis, severe substance use, or acute suicidality) reduces treatment engagement and response
• Chronic hypnotic use (>6 months): Associated with slower improvement trajectories and higher relapse rates, though CBT-I remains effective even in this population with extended treatment

Long-Term Outcome Data

Longitudinal studies show that untreated chronic insomnia is remarkably persistent. The Penn State cohort study by Fernandez-Mendoza et al. found that approximately 46% of individuals with chronic insomnia at baseline still met criteria at 7.5-year follow-up—and that those with persistent insomnia had a significantly elevated risk of incident hypertension, diabetes, and depression. In contrast, treated insomnia patients (particularly those receiving CBT-I) show remission rates of 60–80% at post-treatment, with 50–70% maintaining remission at 12-month follow-up. Booster sessions can restore gains in the majority of relapsers.

Sleep disturbance intersects with nearly every psychiatric diagnosis, but the prevalence, pattern, and clinical significance vary substantially.

Insomnia Comorbidity Prevalence

• Major Depressive Disorder: 60–90% of MDD patients report insomnia; 15–25% report hypersomnia. Insomnia is the most common presenting complaint in primary care depression.
• Generalized Anxiety Disorder (GAD): 56–75% have clinically significant insomnia. Sleep-onset insomnia predominates, driven by worry and cognitive hyperarousal.
• PTSD: 70–91% report insomnia; 50–70% report trauma-related nightmares. Sleep disturbance in PTSD correlates with overall PTSD severity (r = 0.3–0.5 across studies).
• Substance Use Disorders: 36–72% during active use; 25–60% during early recovery. Insomnia during early alcohol recovery is a robust predictor of relapse (OR ≈ 2.0).
• Schizophrenia spectrum: 30–80% across studies, with circadian misalignment as a prominent feature. Sleep disturbance correlates more strongly with negative symptoms and cognitive impairment than with positive symptoms.
• ADHD: 25–50% of adults with ADHD have comorbid insomnia; delayed sleep-wake phase disorder is also overrepresented. The relationship is bidirectional—sleep deprivation worsens attentional symptoms, while ADHD-related hyperarousal disrupts sleep.

Hypersomnia Comorbidity

• Bipolar depression: 38–78% (vs. 15–25% in unipolar depression)
• Atypical depression: ~40% meet criteria for clinically significant hypersomnia
• Seasonal affective disorder (SAD): Hypersomnia is present in approximately 80% and is a defining feature of the winter depression pattern

The clinical implication of these high comorbidity rates is that sleep disturbance should be systematically assessed in every psychiatric evaluation, using validated instruments rather than informal inquiry, and that treatment planning should explicitly address sleep as an independent target.

Beyond insomnia and hypersomnia, circadian rhythm disruption represents a critical but often overlooked dimension of the sleep-psychiatry interface. The circadian system—governed by the suprachiasmatic nucleus (SCN) of the anterior hypothalamus and entrained primarily by light exposure—regulates not only sleep-wake timing but also neurotransmitter release, cortisol rhythmicity, core body temperature, and immune function.

Circadian disruption is pervasive in psychiatric disorders. Delayed sleep-wake phase disorder (DSWPD)—a persistent mismatch between the endogenous circadian clock and desired sleep-wake times—is present in approximately 7–16% of psychiatric outpatients versus 1–4% of the general population. It is particularly common in depression, bipolar disorder, and ADHD, and its misdiagnosis as insomnia or hypersomnia leads to inappropriate treatment (e.g., hypnotic prescriptions that do not address the underlying chronobiological mechanism).

In bipolar disorder, the social zeitgeber theory (Ehlers, Frank, & Kupfer) posits that disruption of social rhythms—regular daily routines that entrain circadian timing—triggers mood episodes. This theoretical framework underlies Interpersonal and Social Rhythm Therapy (IPSRT), which demonstrated efficacy in the Pittsburgh Maintenance Therapies in Recurrent Depression trial, showing longer time to recurrence compared to intensive clinical management.

Emerging evidence suggests that circadian biomarkers—such as dim-light melatonin onset (DLMO), rest-activity rhythm amplitude, and clock gene expression patterns—may serve as objective predictors of treatment response and relapse risk, though clinical application remains largely in the research domain.

Digital and Scalable CBT-I

The most pressing implementation challenge for CBT-I is access. Despite Level I evidence and guideline endorsement, fewer than 1% of insomnia patients receive CBT-I, primarily due to a critical shortage of trained providers. Digital CBT-I (dCBT-I) platforms such as Sleepio, Somryst (now Pear-004, the first FDA-cleared prescription digital therapeutic for insomnia), and CBT-i Coach offer scalable alternatives. Meta-analytic evidence for dCBT-I shows moderate-to-large effects (Hedges' g = 0.40–0.80 on ISI), somewhat smaller than face-to-face CBT-I but clinically meaningful. Completion rates remain a challenge, with approximately 50–65% of users completing full programs.

Sleep-Targeted Suicide Prevention

Given the robust independent association between insomnia and suicidality, an emerging research program is investigating whether insomnia treatment can reduce suicidal ideation. Preliminary RCTs, including the DOZE study and work by Trockel and colleagues, suggest that CBT-I produces reductions in suicidal ideation scores, though large-scale confirmatory trials are ongoing. This represents a potentially transformative application—identifying and treating a modifiable risk factor for suicide.

Glymphatic Function and Neurodegeneration

The discovery of the glymphatic system—a brain waste clearance pathway most active during NREM slow-wave sleep—has opened new avenues connecting sleep disturbance to neurodegeneration. Chronic insomnia with short sleep duration is associated with elevated amyloid-beta and tau accumulation, both hallmarks of Alzheimer's disease. This raises the possibility that sleep disturbance treatment could serve a neuroprotective function, though this remains speculative at present.

Limitations of the Current Evidence Base

• Underrepresentation in trials: Most CBT-I and pharmacotherapy trials have underrepresented racial and ethnic minorities, older adults with multiple comorbidities, and individuals with severe psychiatric illness
• Measurement heterogeneity: Studies vary in their use of subjective versus objective sleep measures, making cross-study comparison difficult
• Hypersomnia treatment evidence is thin: Compared to the robust literature on insomnia, evidence-based treatment options for psychiatric hypersomnia are far less developed, with most pharmacological agents used off-label based on limited trial data
• Long-term pharmacotherapy data: Few hypnotic medication trials extend beyond 3–6 months, creating uncertainty about long-term efficacy and safety for a chronic condition
• Mechanistic understanding gaps: While the hyperarousal model of insomnia is well-supported, the neurobiological mechanisms of psychiatric hypersomnia—particularly why some depressed patients sleep excessively while others become insomnic—remain incompletely understood

The evidence reviewed here supports several key clinical principles for practitioners working at the intersection of sleep and psychiatric medicine:

• Screen systematically: Use validated instruments (ISI, ESS, PSQI) at intake and longitudinally. Sleep disturbance is a vital sign for psychiatric stability.
• Treat insomnia independently: Do not assume that insomnia will resolve with treatment of the "primary" psychiatric condition. It frequently does not, and residual insomnia predicts relapse. The Manber et al. (2008) data—showing doubled depression remission when CBT-I was added to antidepressants—should be practice-changing.
• Prioritize CBT-I: It is the most effective and durable intervention for chronic insomnia, superior to pharmacotherapy on long-term outcomes, and effective across psychiatric comorbidities. When face-to-face CBT-I is unavailable, digital platforms represent a reasonable alternative.
• Evaluate hypersomnia carefully: Always exclude OSA, narcolepsy, circadian disorders, and medication effects before attributing hypersomnia solely to psychiatric illness. Hypersomnia in depression signals a more treatment-resistant phenotype that may require augmentation strategies.
• Monitor sleep as a prodromal marker: In bipolar disorder, psychotic disorders, and recurrent depression, acute changes in sleep pattern—particularly sudden-onset insomnia or reduced sleep need—may be the earliest warning sign of impending episode onset, often preceding mood or psychotic symptoms by days to weeks.
• Assess suicidality in the context of sleep: Given the independent association between insomnia and suicidal behavior, deteriorating sleep should heighten clinical vigilance for suicidal risk, and insomnia treatment should be considered a component of safety planning.