Background: despite developments a substantial part of patients with depression will only recover slowly. Light therapy from light boxes has shown antidepressant effects but have several limitations: time consuming, only allowing a fixed spectral distribution, only delivered at a specific time-point, and often with inadequate light intensity delivered at the retina. Therefore, we developed a new dynamic lighting system using light fixtures that are built into the room and can change intensity and spectral distribution of light during the 24-hour day.

Objectives: the objective of this trial is to assess the beneficial and harmful effects of a newly developed dynamic lighting system using Light Emitting Diodes (LED) -light armatures aiming to mimic sunlight, when installed in the patient rooms of a psychiatric inpatient ward, compared with usual care.

Design: the design is a randomised controlled trial with two arms: an active dynamic light trial arm and a usual care arm with blinding of depression outcome, and data analyses. Randomisation will be 1:1.

Inclusion criteria: a current episode of a major depressive episode as part of a unipolar or bipolar disorder. Patients with bipolar depression should be in current and recent (minimum two months before admission) mood stabilising treatment, age > 18 years, informed consent.

Exclusion criteria: severe suicidality, abuse of alcohol and / or drugs, actual psychotic state, Young Mania Rating score above 7 or fulfilling diagnostic criteria for a current hypomanic or manic episode.

Interventions: the experimental intervention is a dynamic LED-light system in 10 separate patient single rooms with three dynamic lamps: a window jamb built-in light panel, two ceiling mounted lamps, and a wall mounted lamp. The usual care is constant standard LED-light.

Primary outcome: score on the Hamilton Depression Rating Scale 6 item version (HAM-D6) scale at week 3

Secondary outcomes: score on the Suicidal Ideation Attribution Scale (SIDAS ) scale at week 3, and score in the Hamilton. Depression Rating Scale 17 item version (HAM-D17) scale at week 3, and score on the World Health Organisation Quality Of Life questionaire abbreviated version (WHOQOL-BREF) at week 3.

Trial size: in total, 150 patients.

Time schedule: the trial will be submitted for regulatory approvals January 2019, the first participant will be included April 2019, the expected last follow-up of the last participant will be December 2020, the expected last follow-up after 6 months will be June 2021, data will be analysed from June 2021 till September 2021, manuscripts will be prepared from December 2020, and we expect to submit first manuscript December 2021.

The participant population: depression is a common disorder with a prevalence of about 150,000 in Denmark. According to WHO depression is a leading cause of disability worldwide and is a major contributor to the global burden of disease. Society suffers large direct and indirect losses in terms of lost work, sickness and early retirements. In Denmark, the costs of depression alone amount to an estimated 14 billion DKK per year. Depression can affect an individual to the extent that everyday activities and chores are insurmountable, and can ultimately lead to despair, hopelessness, and suicidal behaviour.

Current treatment: the current treatment of a depressive episode, whether it is part of a Major Depressive Disorder (MDD) or of a Bipolar Disorder (BD), involves as first line treatment psychotherapy /psychoeducation, mood stabilisers, and antidepressants. Second line treatments include neurostimulation (Electroconvulsive treatment), and complementary options include exercise, chronotherapeutics (light therapy, sleep deprivation), neurostimulation and other therapies. Recent drug development has not been able to produce compounds with better antidepressant effect, and despite the ongoing development of new psychotherapeutic approaches, improvement from depression is often slow and with 10-25% ending up being treatment resistant. Relapse, readmission, and suicide are also major treatment challenges.

Trial interventions: the present efficacy trial investigates the possible antidepressant effect of a new form of dynamic lighting in the treatment of a major depressive episode either as part of a MDD or a BD. The Light-Emitting-Diode (LED) technology provides new potentials to adjust frequency distribution and intensity of light during the 24 hour-day. LED-lighting can be adjusted to supply light rich in the blue, short waved region of the spectral composition in the first part of the day, and warmer light with less blue later in the day and in the evening. This regulation should, according to our current chronobiological knowledge, entrain and regulate the sleep-wake cycle, and thereby improve and stabilize mood, providing an added non-pharmacological treatment of in-patients suffering from depression.

The experimental intervention specifically consists of implementation of a dynamic LED-light system in patient rooms with three modules:

• a dynamic "sunlight therapy LED-lighting" that mimics the spectral composition and intensity of daylight coming through a southeast (SE) facing window at equinox with 12 hours of sunlight and 12 hours of darkness. The fixture is built into the window jamb (vertical side of the window) and cannot be switched off.
• Two LED-lighting lamps mounted horizontally in the ceiling providing dynamic spectral composition and intensity dynamic. Can be switched on/off.
• A reading LED-light mounted vertically on the wall adjacent to the bed providing dynamic spectral composition and intensity. Can be switched on/off.

The usual care intervention is standard, constant LED-Light with two elements:

• Two LED-Lighting lamps mounted horizontally in the ceiling. Spectral composition and intensity are constant. The ceiling lamps can be switched on/off.
• A reading LED-lighting lamp mounted vertically on the wall adjacent to the bed. Spectral composition and intensity are constant. The reading lamp can be switched on/off.

Bright Light Therapy: the effect of light therapy in the treatment of non-seasonal depression has been studied for several decades. In 2004, Tuunainen and colleagues published the Cochrane systematic review "Light therapy for non-seasonal depression". The conclusion was that light therapy must be regarded as a promising treatment method, but because of the heterogeneity among the studies, methodological problems and a lack of systematic collection of adverse events (AE's), the recommendation of light therapy as treatment of depression should be considered with some caution. In 2007, Even et al found, in a systematic review, an added antidepressant effect of light therapy when used as augmentation to antidepressant therapy. In 2016, Perara et al in a systematic review including 20 RCT's using light therapy for non-seasonal depression. found an overall small antidepressant effect (SMD -0.41; 95% CI -0.64 to -0.18), but with a high risk of bias and inconsistency between studies. None of the included subgroup analyses were significant (stand-alone light therapy versus adjunctive light therapy; morning light therapy versus evening light therapy or other times of day; light therapy for in- versus outpatients; placebo conditions using some form of light versus using non-light-based placebo). However, when analysed separately some support was found for better effect of light when used as monotherapy; in the morning; for outpatients, and when using non-light-based placebos. Only four of the 20 studies had low risk of bias on all items on the Cochrane Risk of Bias Tool. The Martiny et al study found a statistically significant better effect of light therapy plus sertraline compared to placebo light and light therapy, in 102 outpatients. The Lam et al study compared, in 122 outpatients in four groups: (a) active light plus active fluoxetine, (b) active light plus placebo fluoxetine, (c) inactive negative ion generator plus active fluoxetine, (d) inactive negative ion generator plus placebo fluoxetine. They found a statistically significant better effect in the group receiving the combination of active light plus active fluoxetine, and in the group receiving the combination of active light and placebo fluoxetine compared to the group receiving the combination of inactive negative ion generator plus placebo fluoxetine. The Lieversee et al study found, in 84 elderly (> 60 years) outpatients with non-seasonal depression, a statistically significant better effect of active versus placebo light treatment. The Loving et al study found, in 81 elderly outpatients (> 60 years) with non-seasonal depression, treated with bright or placebo light (administered at three different time-points), no difference between groups and a low reduction in depression severity across groups, of 16 %. A more recent study showed effect of bright light treatment (BLT) compared to dim light treatment, in 46 patients with bipolar depression, when used as augmentation in patients receiving mood stabilising medication. A recent review, from Benedetti et al, showed that the risk of switch from depression to mania, in BP patient treated with light therapy, was no higher than with antidepressant drug therapy. With conventional light therapy, patients are placed in front of a light box for 30-60 minutes in the morning. With dynamic lighting, LED-light is built into the room. This enables the system to provide light in the daytime that can phase advance the circadian rhythms and provide an alerting effect. During evening and night, the light system provides low intensity light with low blue wavelength content with minimal impact on the circadian system. Dynamic lighting has, in some studies, been found to provide more stable mood and better sleep quality in people suffering from dementia and a more stable circadian rhythm in patient suffering from depression.

Architecture at Hospitals: the discovery of the non-visual, light sensitive ganglionic cells in the human retina (the intrinsically photosensitive Retinal Ganglion Cells, ipRGCs), with a maximum sensitivity for blue light (460 nm - 480 nm), has stimulated research on the effect of using light with different spectral composition in humans. The ipRGC has been shown to function as the prime connection to the suprachiasmatic nuclei (SCN), the master clock in the hypothalamus, and to a range of brain structures known to be involved in depression. The impact of the signals from the ipRGC is thought to be responsible for the ability of light to time and stabilise the sleep-wake cycle, and to adjust the seasonal regulation of serotonin. LED-light can be particularly rich in the blue wavelength spectrum in contrast to conventional compact fluorescent light (CFL). Thus, LED-light can be tuned to impact the regulation of the sleep-wake cycle. The dynamic steering of LED-light is important as blue light in the evening will delay the sleep-wake cycle, which may cause difficulty falling in sleep (sleep onset insomnia). Blue light in the morning, on the other hand, will advance sleep and increase alertness, and as patients with depression tend to be late chronotypes, this should provide a better sleep regulation. Nonresponse in depression has also been linked to late chronotype, and blue light should help to correct the drift in the sleep-wake cycle that patients with depression often experience.

Recent studies confirm ancient architectural principles about exposure to morning sun, such as the late 19th century Nightingale pavilions facing SE aiming at optimising exposure to the morning sun in the darkest winter season. In a previous pilot project, we documented extreme differences in daylight exposure between hospital wards facing SE and hospital wards facing northwest (NW). Measured on a clear day, these differences were 57.000 Lux at summer solstice, 38.000 Lux at autumn equinox and 19.000 Lux at winter solstice. Also, the spectral composition of the morning light was richer in the blue region towards SE.

At psychiatric units in hospitals there will always be "a dark side of the building" where facades will receive very little - if any - morning sun during wintertime. Dynamic LED-lighting can be an effective alternative to natural sunlight. Today, there is not enough knowledge on how LED-light can be implemented and used or about the potentially beneficial effects of LED-light on patients with depression. We lack practical research knowledge in this field, which combines medical science, architecture, and engineering.

Clinical data on the control intervention: the control intervention as applied in the usual care arm consists of standard hospital LED lighting fixtures without dynamic or spectral changes. The lighting setup in the control intervention consists of two ceiling lamps and one reading lamp, all with LED standard technology.

Handling risks: LED-lights and armatures are approved for use in the psychiatric ward.

Expected side effect: most common side effects are headache, eyestrain, nausea, and agitation. Evening administrations of very bright light have been shown to be associated with sleep disturbances and bipolar patients may switch into hypomania during BLT. The risk of hypomania is, however, considered small, as all patients will be in mood stabilizing drug therapy, and the risk of switch from depression to mania, with light therapy, has been found to be small.

Expected benefits: we expect the dynamic LED-lighting system to provide a larger antidepressant effect than the standard LED-light, better sleep and less use of medication.

Ethical justification and trial rationale: it is in the public and in the patients shared interest to investigate how and if light administration in patient rooms has a potential in the treatment of depression. The intervention in not expected to expose participants to unnecessary risks, because the treatment in the ward will continue as usual and there will be close monitoring as the intervention runs. The potential side effects of dynamic LED-light are expected to be mild and rare. The risk of switch from depression to manic state, due to light therapy is considered small because all participants will receive adequate mood stabilizing treatment, and because the risk of switch to mania from depression with light therapy has been shown to be small. We will exclude all patients with manic symptoms. To find any emergent manic symptoms during the trial, we will rate weekly for manic symptoms and alert the staff to look out for emergence of manic symptoms.

Inclusion Criteria:

• A current episode of a major depressive episode (DSM-5) as part of a unipolar or bipolar disorder.
• Patients with bipolar depression should be in current and recent (minimum two months before admission) mood stabilizing treatment
• Age > 18 years
• Informed consent
• Danish speaking.

Exclusion Criteria:

• Severe suicidality
• Abuse of alcohol and / or drugs
• Actual psychotic state within last four weeks
• Actual Young Mania Rating score above 7 or fulfilling diagnostic criteria for a current hypomanic or manic episode
• Patient is subject to coercive measures of any kind.

We will make data from this study available on reasonable request once the results are published

Blinding of depression severity assessment