Circadian Rhythms and Mental Health: Sleep-Wake Disruption, Light Therapy, Chronotherapy, and Their Role in Mood Disorders

The relationship between circadian rhythms and mental health extends far beyond the commonsense observation that depressed people sleep poorly. Circadian disruption is now understood as a core pathophysiological feature of multiple psychiatric disorders — not merely an epiphenomenon or secondary symptom. The master circadian pacemaker, the suprachiasmatic nucleus (SCN) of the anterior hypothalamus, orchestrates approximately 24-hour oscillations in virtually every physiological system: sleep-wake cycling, core body temperature, cortisol secretion, melatonin production, neurotransmitter turnover, immune function, and gene expression. When this system is disrupted — through genetic vulnerability, environmental misalignment, or disease processes — the consequences for mental health are profound and measurable.

Epidemiological data consistently demonstrate that circadian rhythm sleep-wake disorders (CRSWDs) are substantially more prevalent among individuals with psychiatric conditions than in the general population. In major depressive disorder (MDD), estimates suggest that 75–90% of patients exhibit some form of sleep-wake disturbance, with a significant subset showing objectively measurable circadian phase abnormalities. Bipolar disorder is perhaps the psychiatric condition most tightly linked to circadian disruption: manic episodes are frequently precipitated by sleep deprivation, and the characteristic cycling of mood states maps onto disturbances in circadian periodicity. Seasonal affective disorder (SAD), classified in DSM-5-TR as Major Depressive Disorder with Seasonal Pattern, represents the most direct clinical manifestation of circadian-environmental mismatch, with prevalence rates ranging from 1–10% depending on latitude and methodology.

This article provides an in-depth clinical review of the neurobiological mechanisms linking circadian rhythms to psychiatric illness, the diagnostic landscape of circadian rhythm sleep-wake disorders, the evidence base for chronotherapeutic interventions — including bright light therapy (BLT), sleep deprivation (wake therapy), melatonin and melatonin agonists, and social rhythm therapy — and the prognostic factors that predict treatment response. The aim is to equip clinicians and advanced students with the mechanistic understanding and outcome data necessary to integrate circadian-based approaches into psychiatric care.

Circadian rhythmicity is generated at the molecular level by transcription-translation feedback loops (TTFLs) operating within individual cells. The core loop involves the transcription factors CLOCK and BMAL1, which heterodimerize and drive expression of the Period (PER1, PER2, PER3) and Cryptochrome (CRY1, CRY2) genes. The PER and CRY proteins accumulate in the cytoplasm, form complexes, translocate back into the nucleus, and inhibit CLOCK-BMAL1 activity — thereby repressing their own transcription. This cycle takes approximately 24 hours and is modulated by accessory loops involving REV-ERBα/β and RORα, which regulate BMAL1 expression. Casein kinase 1δ/ε (CK1δ/ε) phosphorylates PER proteins, regulating their stability and the period length of the oscillation. Mutations in these clock genes have direct psychiatric relevance: a missense mutation in CK1δ (T44A) has been identified in familial advanced sleep phase syndrome (FASPS), and polymorphisms in CLOCK, PER2, PER3, and CRY1 have been associated with vulnerability to MDD, bipolar disorder, and SAD in genome-wide association studies (GWAS) and candidate gene studies.

The SCN receives direct photic input via the retinohypothalamic tract (RHT) from intrinsically photosensitive retinal ganglion cells (ipRGCs) containing the photopigment melanopsin, which is maximally sensitive to short-wavelength blue light (~480 nm). This pathway is the primary mechanism of photoentrainment — the synchronization of the internal clock to the external light-dark cycle. The SCN then coordinates peripheral oscillators throughout the body via both neural and humoral outputs. Key efferent pathways include:

• SCN → Paraventricular nucleus (PVN) → Intermediolateral cell column → Superior cervical ganglion → Pineal gland: This multisynaptic pathway regulates melatonin synthesis. SCN output inhibits melatonin production during the day; removal of this inhibition at night triggers melatonin synthesis from serotonin via N-acetyltransferase (AANAT) and hydroxyindole-O-methyltransferase (HIOMT). Melatonin provides a humoral "darkness signal" that feeds back to MT1 and MT2 receptors on the SCN itself, as well as acting on receptors throughout the brain and periphery.
• SCN → Hypothalamic arousal systems: The SCN modulates the activity of orexin/hypocretin neurons in the lateral hypothalamus, histaminergic neurons in the tuberomammillary nucleus (TMN), and the ventrolateral preoptic area (VLPO), which is the key sleep-promoting GABAergic/galaninergic cell group. This connectivity means that circadian disruption directly impacts the "flip-flop switch" model of sleep-wake regulation described by Saper and colleagues.
• SCN → Monoaminergic systems: The SCN influences serotonergic neurons in the dorsal raphe nucleus, noradrenergic neurons in the locus coeruleus (LC), and dopaminergic neurons in the ventral tegmental area (VTA) — all systems centrally implicated in mood regulation. Circadian variation in serotonin turnover, norepinephrine release, and dopamine receptor sensitivity provides a direct neurochemical bridge between clock function and affective state.
• SCN → HPA axis: The SCN regulates cortisol secretion via the PVN-CRH pathway. The characteristic cortisol rhythm — with a nadir around midnight and a peak shortly after awakening (the cortisol awakening response, CAR) — is disrupted in MDD, with depressed patients often showing elevated nocturnal cortisol, flattened diurnal amplitude, and an elevated cortisol nadir. This blunted rhythm has been associated with hippocampal volume reductions and cognitive impairment.

Beyond the SCN, peripheral clocks operate in virtually every tissue, including the prefrontal cortex, hippocampus, amygdala, and striatum. Postmortem studies from the Stanley Neuropathology Consortium and BrainSpan datasets have revealed that the rhythmic expression of clock genes is disrupted in the dorsolateral prefrontal cortex of individuals with MDD and bipolar disorder, with a loss of normal phase relationships between gene expression rhythms. A landmark 2013 study by Li et al. in Proceedings of the National Academy of Sciences examined gene expression in 12,000 transcripts across six brain regions in 55 individuals with MDD and matched controls and found widespread disruption of circadian gene expression patterns, with many genes showing shifted or abolished rhythmicity. This finding suggests that circadian disruption in mood disorders is not limited to the SCN but is a brain-wide phenomenon.

The DSM-5-TR classifies Circadian Rhythm Sleep-Wake Disorders (CRSWDs) under Sleep-Wake Disorders, specifying five subtypes: Delayed Sleep-Wake Phase Type, Advanced Sleep-Wake Phase Type, Irregular Sleep-Wake Type, Non-24-Hour Sleep-Wake Type, and Shift Work Type. The ICD-11 uses largely parallel categories under "Disorders of the Sleep-Wake Schedule" (7A60–7A6Z). Both systems require that the circadian misalignment causes clinically significant distress or functional impairment and is not better explained by another sleep disorder, substance use, or medical condition.

Delayed Sleep-Wake Phase Disorder (DSWPD) is characterized by a habitual sleep-wake timing that is delayed by 2 or more hours relative to socially conventional or desired times, with an inability to advance the sleep phase. Prevalence in the general population is estimated at 0.13–0.17% by strict diagnostic criteria, though the prevalence of the broader phenotype ("evening chronotype") is far higher — approximately 10–20% of the population identifies as a strong evening type on the Morningness-Eveningness Questionnaire (MEQ). In adolescents and young adults, prevalence of DSWPD is substantially higher, estimated at 7–16%. DSWPD has a strong genetic basis, with polymorphisms in PER3 (a variable-number tandem repeat), CRY1 (a gain-of-function variant identified by Patke et al. in 2017), and CK1ε implicated in delayed phase phenotypes.

Advanced Sleep-Wake Phase Disorder (ASWPD) is rarer, with a prevalence estimated at <1% in the general population, and is more common in older adults. Familial ASWPD has been linked to a serine-to-glycine mutation in PER2 (S662G) affecting the CK1ε phosphorylation site, leading to reduced PER2 stability and a shortened circadian period.

Key diagnostic pitfalls include:

• Differentiating DSWPD from insomnia disorder: Patients with DSWPD often present with "sleep-onset insomnia" when attempting to sleep at conventional times, but will show normal sleep architecture and duration when allowed to sleep on their preferred schedule. Actigraphy over 7–14 days (with sleep diary) is the gold standard for distinguishing these conditions. Polysomnography is typically normal in DSWPD.
• Circadian disruption vs. behavioral sleep restriction: Chronic insufficient sleep due to social obligations in DSWPD patients can produce symptoms indistinguishable from primary sleep deprivation — fatigue, cognitive impairment, irritability — which may be misattributed to depression or ADHD. The phenomenon of "social jet lag" (the discrepancy between biological and social clocks) affects an estimated 69% of the population to some degree, with greater misalignment associated with higher BMI, depressive symptoms, and cardiometabolic risk.
• Comorbidity with psychiatric disorders: DSWPD is overrepresented in ADHD (estimated at 73–78% of adults with ADHD showing delayed circadian phase by actigraphy or dim-light melatonin onset [DLMO] assessment), major depression (estimated 10–30% of MDD patients show circadian phase delay), and autism spectrum disorder. These comorbidities complicate differential diagnosis and raise the question of shared genetic architecture.
• Dim-light melatonin onset (DLMO): DLMO is the most reliable circadian phase marker available in clinical practice. In DSWPD, DLMO typically occurs after 21:30–22:00, often as late as 01:00–04:00. In ASWPD, DLMO occurs before 18:30. DLMO measurement requires salivary or plasma melatonin sampling under dim-light conditions (<30 lux) at 30–60 minute intervals across the evening. While not routinely available in clinical practice, it remains the reference standard for circadian phase assessment and is increasingly accessible through specialized sleep labs and home collection kits.

The intersection of circadian disruption and psychiatric illness is not limited to sleep disorders per se. Circadian abnormalities represent a transdiagnostic feature cutting across mood, psychotic, neurodevelopmental, and anxiety disorders, with distinct patterns for each diagnostic category.

Major Depressive Disorder

Sleep disturbance — either insomnia or hypersomnia — is one of the nine cardinal criteria for MDD in DSM-5-TR and is present in approximately 80–90% of depressed patients. Beyond subjective complaints, objective circadian abnormalities in MDD include: shortened REM latency (the interval from sleep onset to the first REM period, which normally averages ~90 minutes but may be reduced to 40–60 minutes in MDD), increased REM density (the frequency of rapid eye movements during REM), phase-advanced or phase-delayed circadian rhythms depending on the depressive subtype, blunted circadian amplitude of core body temperature and cortisol, and altered melatonin profiles. The two-process model of sleep regulation (Process S: homeostatic sleep drive; Process C: circadian alerting signal) proposes that in depression there is a mismatch between these processes — specifically, that the circadian system may be phase-advanced relative to the sleep-wake cycle, leading to early REM onset and mood disturbance concentrated in the early morning hours. This "internal desynchronization" hypothesis, elaborated by Wehr and Wirz-Justice in the 1980s, was a foundational framework for chronotherapeutic approaches.

The melancholic subtype of MDD, characterized by early morning awakening, diurnal mood variation (worse in the morning), and psychomotor disturbance, shows the most consistent circadian phase advancement and cortisol rhythm abnormalities. In contrast, atypical depression — with hypersomnia, leaden paralysis, and mood reactivity — may show phase delay and increased sleep duration. This phenotypic distinction has therapeutic implications: melancholic features may predict better response to wake therapy and morning bright light, while atypical features may respond preferentially to evening light restriction and strategic melatonin administration.

Bipolar Disorder

Bipolar disorder has the strongest bidirectional relationship with circadian disruption of any psychiatric condition. Sleep deprivation can trigger manic episodes — a finding replicated across multiple studies and clinically well-recognized. Conversely, mania is characterized by markedly reduced sleep need (a DSM-5-TR criterion), often dropping to 2–4 hours per night without subjective fatigue. Bipolar depression frequently involves hypersomnia (in 38–78% of bipolar depressive episodes, depending on the study) and delayed circadian phase.

At the molecular level, genome-wide association studies have identified associations between bipolar disorder and variants in clock genes including CLOCK, ARNTL (BMAL1), PER3, TIMELESS, and REV-ERBα (NR1D1). The CLOCK 3111T/C polymorphism has been associated with evening chronotype and mood instability in bipolar patients. Animal models further support this link: Clock-mutant mice exhibit a behavioral phenotype that closely resembles mania — hyperactivity, reduced sleep, impulsivity, increased reward-seeking, and elevated dopaminergic tone in the VTA — which is reversible with lithium administration. Notably, lithium directly inhibits glycogen synthase kinase 3β (GSK-3β), an enzyme that phosphorylates and stabilizes REV-ERBα, thereby modulating the molecular clock. This mechanism may partly explain lithium's mood-stabilizing effects and its ability to lengthen the circadian period.

Seasonal Affective Disorder (MDD with Seasonal Pattern)

SAD affects an estimated 1–3% of the general population in temperate climates, with prevalence increasing dramatically with latitude: approximately 1.4% at 27°N (Florida) versus 9.7% at 64°N (Alaska), based on the landmark epidemiological surveys by Rosen et al. (1990). The phase-shift hypothesis, proposed by Lewy and colleagues, posits that SAD results from a circadian phase delay relative to the sleep-wake cycle during the short photoperiods of winter. Supporting evidence includes the finding that DLMO is delayed in SAD patients during winter relative to summer, and that morning bright light therapy (which phase-advances the clock) is more effective than evening light — suggesting that correcting the phase delay is therapeutically critical.

Schizophrenia Spectrum Disorders

Circadian disruption in schizophrenia is pervasive but underrecognized. Actigraphic studies suggest that 44–80% of individuals with schizophrenia have disrupted rest-activity rhythms, with some meeting criteria for irregular sleep-wake rhythm disorder or non-24-hour sleep-wake disorder. Postmortem studies have shown disrupted clock gene expression in the prefrontal cortex and thalamus of individuals with schizophrenia. Antipsychotic medications, particularly those with high histamine H1 and serotonin 5-HT2C receptor affinity (e.g., olanzapine, quetiapine), can further disrupt circadian rhythms and metabolic cycling.

ADHD

An estimated 73–78% of adults with ADHD demonstrate delayed circadian phase, with delayed DLMO and later sleep-wake timing. This has led some researchers to propose that ADHD and DSWPD share a common circadian etiopathology. Stimulant medications (methylphenidate, amphetamines) can further delay circadian phase through dopaminergic effects on the SCN and masking of sleepiness, potentially perpetuating the cycle. The International Consensus Statement on ADHD and Sleep (2019) recommended routine circadian assessment in ADHD management.

Bright light therapy (BLT) is the best-studied chronotherapeutic intervention and the first-line treatment for SAD. The mechanism of action involves stimulation of ipRGCs in the retina, which project via the RHT to the SCN, where they modulate clock gene expression and phase-shift the circadian pacemaker. Morning light exposure phase-advances the clock (useful in DSWPD and winter SAD), while evening light phase-delays it (useful in ASWPD). Beyond phase-shifting, BLT also has direct alerting effects mediated by suppression of melatonin, stimulation of cortisol release, and enhancement of serotonergic transmission — effects that may explain its antidepressant properties independent of circadian phase correction.

Evidence in Seasonal Affective Disorder

The efficacy of BLT in SAD is supported by over 60 randomized controlled trials and multiple meta-analyses. The landmark meta-analysis by Golden et al. (2005), commissioned by the American Psychiatric Association, found that BLT for SAD had an effect size (Cohen's d) of 0.84 — comparable to antidepressant medication for non-seasonal depression. Response rates (≥50% reduction in depression rating scale scores) in RCTs typically range from 53–80% with BLT versus 30–38% with placebo (dim light or deactivated ion generator). Remission rates range from 40–60%. The NNT for BLT in SAD, derived from pooled RCT data, is approximately 3–4 — notably better than the NNT of 5–9 for SSRIs in moderate depression.

Standard parameters for BLT in SAD include: 10,000 lux white fluorescent light (full-spectrum or enriched in blue wavelengths), positioned at eye level at approximately 30–40 cm distance, for 30 minutes daily, initiated within 30 minutes of habitual wake time. Lower intensities (2,500 lux) require proportionally longer exposure times (~2 hours). Clinical response typically begins within 3–5 days, with full response by 2–4 weeks. The CAN-SAD trial (Lam et al., 2006), a landmark 8-week double-blind RCT comparing BLT (10,000 lux, 30 min/day) versus fluoxetine 20 mg/day versus placebo, found no significant difference between BLT and fluoxetine on the primary outcome (HAM-D scores), with both superior to placebo. BLT showed a trend toward earlier onset of response (week 1 vs. week 2).

Evidence in Non-Seasonal Depression

A critical and underappreciated body of evidence supports BLT for non-seasonal MDD. The meta-analysis by Perera et al. (2016) in JAMA Psychiatry, pooling 20 RCTs (n = 620), found that BLT as monotherapy or as adjunct to antidepressants was effective for non-seasonal depression, with an overall effect size of d = 0.41 (95% CI: 0.18–0.63) as monotherapy and d = 0.63 (95% CI: 0.28–0.98) as an adjunct. Response rates ranged from 44–67% for BLT versus 19–37% for control conditions. Particularly notable was the RCT by Lam et al. (2016) in JAMA Psychiatry, which found that BLT monotherapy was superior to fluoxetine 20 mg in non-seasonal MDD over 8 weeks (response: 50.0% vs. 28.6%; remission: 43.8% vs. 19.4%), with the combination of BLT + fluoxetine showing the highest response (75.9%) and remission (58.6%) rates.

Evidence in Bipolar Depression

BLT is an emerging treatment for bipolar depression, where antidepressant medication carries the risk of mood switching. A 2018 RCT by Sit et al. in the American Journal of Psychiatry found that midday bright white light (7,000 lux, 15–60 minutes, titrated) produced a remission rate of 68.2% versus 22.2% for dim red placebo light in bipolar depression at 6 weeks. Critically, no manic switches were observed with midday timing, addressing a key clinical concern. Midday timing (rather than morning) was chosen to minimize the risk of phase-advancing into mania.

Side Effects and Contraindications

BLT is generally well-tolerated. The most common side effects include headache (10–20%), eye strain (5–15%), nausea (5–10%), and agitation or irritability (5–10%), which are typically transient and dose-dependent. The risk of hypomania or mania induction is a genuine concern in bipolar patients receiving morning BLT and underscores the importance of mood monitoring and midday timing in this population. Relative contraindications include pre-existing retinal disease, photosensitizing medications (lithium, certain antipsychotics, tetracyclines), and a history of rapid cycling bipolar disorder. There is no evidence that BLT at therapeutic intensities causes retinal damage in individuals with healthy eyes.

Total sleep deprivation (TSD), also termed wake therapy, is one of the most rapid-acting antidepressant interventions known — and paradoxically, one of the least utilized. The observation that one night of complete sleep deprivation can produce dramatic mood improvement in depressed patients was first reported by Pflug and Tölle in 1971, and has been replicated extensively. Meta-analyses estimate that 40–60% of MDD patients show a clinically significant antidepressant response to a single night of TSD (defined as sustained wakefulness from morning to the following evening, approximately 36 hours). This response is rapid — often emerging by the early morning hours — and large in magnitude, with effect sizes comparable to or exceeding those of conventional antidepressants.

The principal limitation is that the antidepressant effect of TSD is transient: approximately 50–80% of responders relapse after recovery sleep, even a brief nap. This has led to the development of chronotherapeutic combination protocols — typically referred to as "triple chronotherapy" — which combine TSD with sleep phase advance (SPA) and morning BLT to sustain the antidepressant response. In these protocols, after one night of TSD, the patient is allowed a shortened sleep period (e.g., 21:00–01:00 on night 2, 22:00–02:00 on night 3, gradually advancing to normal timing over 5–7 days) while receiving morning BLT. Some protocols also add lithium or an SSRI.

The landmark Colombo chronotherapy protocol, developed by Benedetti, Colombo, and colleagues at San Raffaele Hospital in Milan, combined repeated TSD (3 sessions over 1 week), SPA, and BLT with ongoing lithium therapy in bipolar depression. Their results showed sustained response rates of approximately 70% at one week and 55–60% at 3 months — remarkably high for bipolar depression, where antidepressant response rates to pharmacotherapy alone are often 30–40%. A 2014 systematic review by Echizenya et al. confirmed that combining TSD with SPA and/or BLT substantially reduces relapse rates compared to TSD alone.

The mechanism of wake therapy remains incompletely understood but involves several converging pathways. Sleep deprivation increases synaptic adenosine (a homeostatic sleep factor that accumulates during wakefulness), which enhances adenosine A1 receptor-mediated inhibition in the prefrontal cortex and may "reset" cortical excitability. TSD also rapidly increases brain-derived neurotrophic factor (BDNF), enhances dopaminergic and serotonergic transmission, and reduces amygdala reactivity to negative emotional stimuli — effects that parallel those of ketamine and electroconvulsive therapy (ECT). Neuroimaging studies show that responders to TSD exhibit baseline hyperactivation of the anterior cingulate cortex (ACC) and medial prefrontal cortex that normalizes after sleep deprivation.

Despite its strong evidence base, wake therapy is rarely used in clinical practice outside specialized European centers. Barriers include the need for overnight monitoring, patient discomfort, the perception of impracticality, and unfamiliarity among clinicians. However, inpatient psychiatric settings are well-suited for its implementation, and emerging outpatient protocols using modified partial sleep deprivation (late-night or second-half-of-night deprivation) show promise with somewhat lower but still meaningful response rates (30–40%).

Exogenous melatonin and melatonin receptor agonists represent a pharmacological approach to circadian rhythm correction. Melatonin acts primarily at MT1 receptors (mediating sleep promotion via SCN inhibition) and MT2 receptors (mediating circadian phase-shifting). Its chronobiotic properties — the ability to shift the timing of the circadian clock — are distinct from its mild soporific effects. The phase-response curve (PRC) for melatonin is approximately opposite to that of light: melatonin administered in the early evening (5–7 hours before habitual DLMO) produces a phase advance, while morning administration produces a phase delay.

For DSWPD: The American Academy of Sleep Medicine (AASM) clinical practice guidelines (2015) recommend low-dose melatonin (0.5–3 mg), administered approximately 5 hours before desired sleep time, for DSWPD. A meta-analysis by van Geijlswijk et al. (2010) found that melatonin advanced sleep onset by an average of 34 minutes and advanced DLMO by 1.18 hours. The quality of evidence was rated as moderate. Importantly, higher doses (5–10 mg) are not more effective for phase-shifting and may be less effective because the extended pharmacokinetic profile can produce melatonin exposure at the wrong circadian phase, partially counteracting the desired advance.

Ramelteon (Rozerem) is a selective MT1/MT2 agonist with higher receptor binding affinity than melatonin (Ki: 0.014 nM at MT1 vs. 0.081 nM for melatonin). It is FDA-approved for insomnia characterized by difficulty with sleep onset. Evidence for its efficacy in CRSWDs is limited to small studies, but it has been used off-label for DSWPD with some success.

Tasimelteon (Hetlioz) is an MT1/MT2 agonist FDA-approved specifically for Non-24-Hour Sleep-Wake Disorder in totally blind individuals — a population that lacks photic entrainment and is therefore dependent on non-photic zeitgebers. In the RESET and SET trials, tasimelteon entrained the circadian melatonin and cortisol rhythms in approximately 20% of blind patients (NNT ~5), a modest but clinically meaningful effect in a condition with no other pharmacological treatment.

Agomelatine (Valdoxan) is a unique antidepressant that combines MT1/MT2 agonism with 5-HT2C antagonism. Licensed in Europe and Australia (but not the United States) for MDD, agomelatine has demonstrated antidepressant efficacy comparable to SSRIs and SNRIs in several RCTs, with particular benefits in improving sleep quality and restoring circadian rhythmicity. A meta-analysis by Taylor et al. (2014) found an effect size of d = 0.24 (95% CI: 0.12–0.35) versus placebo for depressive symptom reduction — modest, though comparable to other antidepressants when analyzed by the same stringent methodology. Its chronobiotic properties distinguish it mechanistically from other antidepressants and make it particularly suitable for depressed patients with prominent circadian disruption. Liver function monitoring is required due to a small risk of hepatotoxicity (elevated transaminases in ~1% of patients).

Social and behavioral zeitgebers — mealtimes, exercise, social interactions, work schedules — are powerful non-photic entrainers of the circadian system. The disruption of these routines (termed "social rhythm disruption") has been shown to precipitate mood episodes in vulnerable individuals. Ehlers, Frank, and Kupfer developed the Social Zeitgeber Theory of mood disorders, which posits that life events disrupt social routines, leading to circadian instability, which in turn triggers depressive or manic episodes in biologically predisposed individuals.

This theory provided the foundation for Interpersonal and Social Rhythm Therapy (IPSRT), developed by Ellen Frank and colleagues at the University of Pittsburgh. IPSRT combines elements of interpersonal therapy (IPT) — addressing grief, role disputes, role transitions, and interpersonal deficits — with a structured behavioral component focused on stabilizing daily routines using the Social Rhythm Metric (SRM), a self-report measure of the regularity of 17 daily activities.

The landmark evidence for IPSRT comes from the Maintenance Therapies in Bipolar Disorder (MTBD) study, a multi-year RCT comparing IPSRT versus intensive clinical management (ICM) in acute and maintenance phases of bipolar I disorder. Key findings included:

• Patients assigned to IPSRT in the acute phase had a longer time to relapse during maintenance (median 67 weeks vs. 40 weeks for ICM, p = 0.01), regardless of maintenance treatment assignment.
• The ability of IPSRT to increase SRM regularity scores during acute treatment predicted longer survival times without new episodes.
• IPSRT did not show a significant advantage over ICM in the time to acute stabilization, suggesting its primary benefit is in relapse prevention rather than acute symptom reduction.

IPSRT is recommended by the CANMAT/ISBD (2018) guidelines as a first-line adjunctive psychotherapy for bipolar disorder maintenance and as a second-line treatment for acute bipolar depression. Its effect sizes for relapse prevention (HR ≈ 0.56 for any mood episode) are comparable to those of maintenance pharmacotherapy.

Beyond IPSRT, other behavioral approaches include:

• Sleep restriction therapy: Adapted from its use in insomnia, controlled sleep restriction can be used to consolidate sleep and enhance circadian amplitude in patients with disrupted or irregular sleep-wake patterns.
• Stimulus control therapy: Standard behavioral insomnia interventions (bed used only for sleep, consistent wake time, avoidance of napping) serve as non-specific circadian stabilizers.
• Scheduled exercise: Moderate-intensity exercise, particularly in the morning, acts as a non-photic zeitgeber and has been shown to phase-advance the circadian clock by up to 1 hour in laboratory studies.
• Dark therapy and blue-light-blocking glasses: Restricting light exposure in the evening — either through enforced darkness or amber/blue-blocking lenses — has shown preliminary efficacy in acute mania. A 2016 RCT by Henriksen et al. found that blue-blocking glasses worn from 18:00 to 08:00 produced a significantly greater decline in Young Mania Rating Scale (YMRS) scores compared to clear-lens controls over 7 days in hospitalized manic patients (effect size d = 1.86).

Identifying patients most likely to benefit from chronotherapy is essential for personalized treatment planning. Several clinical, biological, and genetic factors have been identified that predict response to BLT, wake therapy, and other circadian interventions.

Predictors of BLT Response

• Seasonal pattern: Patients with a clear seasonal pattern (winter depression) show the highest response rates to BLT (53–80%). Non-seasonal depression shows somewhat lower but still clinically significant response rates (44–67%).
• Atypical depressive features: Hypersomnia, carbohydrate craving, and weight gain — features of the atypical subtype and common in SAD — predict better BLT response. This may relate to serotonergic mechanisms, as these features implicate central serotonin deficiency.
• Circadian phase delay: Patients whose DLMO is delayed relative to their habitual sleep time (indicating a phase-delayed clock) show better responses to morning BLT, consistent with the phase-shift hypothesis.
• Diurnal mood variation: Patients with mood that is worst in the morning and improves throughout the day (a melancholic feature) may respond to BLT, though the evidence is mixed.
• Family history of SAD or mood disorders: A positive family history predicts BLT response, suggesting shared genetic susceptibility.

Predictors of Wake Therapy Response

• Diurnal mood variation: Patients with marked diurnal variation (worst in the morning) are more likely to respond to TSD, with response rates as high as 60–67% versus 30–40% in those without diurnal variation.
• Baseline anterior cingulate cortex (ACC) hypermetabolism: PET and fMRI studies have identified elevated pretreatment activity in the ACC (particularly subgenual ACC, Brodmann area 25) as a predictor of TSD response — the same biomarker that predicts response to SSRIs and ECT.
• Higher baseline depression severity: Moderately to severely depressed patients show higher absolute response rates than mildly depressed patients, though the proportional benefit is similar.
• Bipolar vs. unipolar depression: Some studies suggest higher TSD response rates in bipolar depression (50–70%) than in unipolar depression (40–60%), though this finding is not consistent across all meta-analyses.

Genetic Predictors

Emerging pharmacogenomic research has identified circadian gene polymorphisms associated with treatment response. Variation in PER2, GSK-3β, and CLOCK has been associated with lithium response in bipolar disorder. The PER3 variable-number tandem repeat (VNTR) polymorphism — with the longer 5-repeat allele associated with morning preference and the shorter 4-repeat allele with evening preference — may predict differential response to chronotherapy, though findings require replication in larger samples. Pharmacogenomic-guided chronotherapy remains a research frontier rather than a clinical reality.

Direct head-to-head comparisons among chronotherapeutic modalities are limited but informative. The available evidence allows some provisional conclusions about the relative efficacy and positioning of different approaches.

BLT vs. Antidepressant Medication: The CAN-SAD trial (Lam et al., 2006) found no significant difference between BLT and fluoxetine 20 mg for SAD over 8 weeks. The Lam et al. (2016) trial for non-seasonal MDD found BLT was actually superior to fluoxetine as monotherapy at 8 weeks. These findings position BLT as a credible first-line treatment comparable to pharmacotherapy, with the advantages of rapid onset, absence of sexual dysfunction, and no withdrawal effects.

BLT vs. CBT for SAD: A landmark RCT by Rohan et al. (2016) compared BLT versus group CBT-SAD versus their combination across two consecutive winter seasons. In the first winter, BLT and CBT-SAD showed equivalent acute response rates (~58%). However, at long-term follow-up in the second winter, CBT-SAD showed superior outcomes — lower recurrence rates (27.3% vs. 45.6% for BLT) — suggesting that CBT provides durable relapse prevention skills that outlast the acute effects of light exposure.

Wake Therapy vs. Pharmacotherapy: No adequately powered head-to-head RCT of TSD alone versus antidepressant medication has been published. However, the speed of onset of TSD (hours) dramatically outpaces conventional antidepressants (weeks). The Colombo chronotherapy protocol showed response rates (~70%) that approach those of ECT and ketamine for treatment-resistant depression, with a substantially better side effect profile. A 2017 randomized trial by Martiny et al. found that triple chronotherapy (wake therapy + sleep phase advance + BLT) combined with duloxetine produced significantly faster and greater antidepressant response at 1 week (41% remission) compared to duloxetine + exercise control (13% remission), with the advantage sustained at 29 weeks.

Combination Approaches: The evidence consistently supports combination strategies over monotherapy. BLT + antidepressant > BLT alone (Lam et al., 2016). TSD + SPA + BLT >> TSD alone (Colombo protocol). BLT + IPSRT has not been formally compared to either alone in bipolar disorder but represents a logical combination. The emerging consensus is that chronotherapeutic approaches should be conceptualized as a toolkit of complementary interventions rather than competing alternatives.

Despite a compelling body of evidence, the circadian-psychiatric field faces several significant challenges and active research frontiers.

Methodological Limitations

Blinding in chronotherapy trials is inherently difficult. Light therapy trials have used dim-light or inactive negative ion generators as placebos, but complete blinding is rarely achievable. Wake therapy cannot be blinded by definition. This methodological limitation means that effect sizes may be partially inflated by expectancy effects, and the overall evidence quality for chronotherapy, while substantial, is generally rated lower by evidence-grading systems (e.g., GRADE) than comparably effective pharmacological interventions. The field needs creative solutions for sham-controlled designs and larger, multisite RCTs.

Precision Chronotherapy

One of the most promising research frontiers is the development of individualized chronotherapy based on circadian biomarkers. Current approaches include: wearable actigraphy with machine learning algorithms to estimate circadian phase in real time, home DLMO collection kits, and computational models predicting individual phase response curves to light exposure. The goal is to move from population-level treatment protocols ("10,000 lux for 30 minutes in the morning") to individually timed interventions optimized for each patient's circadian phase. Several groups are developing smartphone-based platforms that integrate actigraphy, light exposure data, mood ratings, and predictive algorithms to recommend personalized light timing.

Circadian Biomarkers in Treatment Response Prediction

Research is exploring whether circadian parameters — rest-activity rhythm fragmentation, amplitude, acrophase, and phase angle between DLMO and sleep onset — can serve as predictive biomarkers for treatment response across psychiatric conditions. The UK Biobank actigraphy data (from approximately 100,000 participants wearing wrist accelerometers for 7 days) has enabled population-level analyses showing that lower rest-activity rhythm amplitude and greater fragmentation are associated with MDD, bipolar disorder, and subjective loneliness, independent of total activity levels. Whether these parameters predict treatment response in prospective trials is under active investigation.

Circadian Interventions in Acute Inpatient Settings

Hospitalized psychiatric patients are exposed to chronically disrupted circadian environments — often with low daytime light levels (<100 lux in many hospital settings vs. >10,000 lux outdoors), constant nighttime disruptions, and irregular schedules. Emerging research suggests that architectural and environmental modifications — bright daytime lighting (≥1,000 lux in common areas), evening light restriction, and protected sleep periods — can improve outcomes. A 2017 study by Canazei et al. in a psychiatric inpatient ward found that dynamic lighting systems mimicking natural light patterns reduced length of stay by an average of 2.4 days.

The Gut-Circadian-Mood Axis

An emerging research area is the interaction between the gut microbiome and circadian rhythms. Gut microbiota show diurnal compositional oscillations that are influenced by meal timing and host clock gene expression. Disruption of these microbial rhythms — through circadian disruption, shift work, or jet lag — has been associated with metabolic dysregulation and inflammatory changes that may contribute to mood pathology. This represents a potential mechanistic link between circadian disruption, metabolic comorbidity, and psychiatric illness, though the translational implications remain speculative.

The evidence base for circadian-based approaches in psychiatry has reached a level of maturity that warrants their integration into standard clinical practice. The following practical recommendations synthesize the current evidence:

• Screen all patients with mood disorders for circadian disruption. This includes assessing chronotype (Morningness-Eveningness Questionnaire), sleep-wake regularity (sleep diary, actigraphy if available), seasonal patterns, shift work history, and light exposure habits. DLMO measurement should be considered when circadian phase disorder is suspected.
• BLT should be considered a first-line treatment for SAD (NNT ~3–4), with efficacy comparable to SSRIs and faster onset. It is also a well-supported adjunctive treatment for non-seasonal MDD and an emerging treatment for bipolar depression (using midday timing).
• Wake therapy (TSD) combined with sleep phase advance and BLT is the most rapidly acting non-pharmacological antidepressant intervention available. It is particularly suited for inpatient settings and should be considered for patients requiring urgent mood improvement, including those with suicidal ideation requiring rapid stabilization.
• IPSRT should be offered to patients with bipolar disorder as a maintenance psychotherapy with evidence-based relapse prevention effects comparable to maintenance pharmacotherapy.
• Low-dose melatonin (0.5–3 mg), timed 5 hours before desired sleep onset, is appropriate for DSWPD. Higher doses do not improve phase-shifting efficacy.
• Evening light restriction — including blue-light-blocking glasses, reduced screen use, and dimmer ambient lighting after sunset — is a low-risk, low-cost intervention that may benefit patients with delayed circadian phase, insomnia, or acute mania.
• Stabilize social rhythms in all patients with recurrent mood disorders. Regular sleep-wake times, consistent meal times, and structured daily routines function as behavioral zeitgebers that support circadian stability.
• Consider circadian factors in medication timing: Lithium's chronobiotic effects (via GSK-3β inhibition) and agomelatine's MT1/MT2 agonism represent pharmacological approaches to circadian stabilization that may be particularly suitable for patients with prominent circadian disruption.

The overall trajectory of the field points toward an increasingly sophisticated integration of circadian biology into psychiatric diagnosis and treatment — moving from the current population-level approach toward precision chronotherapy guided by individual circadian biomarkers, genetic profiles, and real-time monitoring. For clinicians, the immediate practical opportunity is to recognize circadian disruption as a modifiable therapeutic target and to deploy the chronotherapeutic toolkit — light, timing, rhythm, and darkness — alongside conventional pharmacological and psychotherapeutic approaches.