Light stimulates the human visual system and the biological functions in the retina, also referred to as non-visual responses, e.g. hormone production. Exposure to the correct light composition can produce acute alertness and increase good sleep quality (1).

In this study, subjects will be exposed to 24-hour LED naturalistic lighting (intervention) or traditional lighting (control) during all days of hospitalization. Subjects will be blinded to the intervention as they will not be told if they are admitted to an intervention patient room or a control patient room.

The primary outcome measure is cortisol levels measured in saliva samples. Secondary outcome measures are delirium rates, length of admission, use of constant observation, mortality and adverse advent.

It is estimated that 80 subjects will be included in the study.

During hospitalization, patient rooms are often associated with poor lighting conditions, and patients are often not exposed to outdoor activities. Many patients have difficulties sleeping during hospital admission, which affects health outcomes and potentially leads to prolonged admission and rehabilitation and a higher risk of developing delirium. However, the circadian rhythm can be modified with LED light, as this technology can reach sufficient levels to affect the human melanopic equivalent daylight illuminance (Melanopic EDI). A high melanopic EDI during the day is supportive for alertness and a good night's sleep. A good night's sleep is essential to prevent the development of delirium. LED lighting, which can give melanopic EDI, is called naturalistic light (1,2).

In this study, subjects will be exposed to 24-hour LED naturalistic lighting (intervention) or traditional lighting of fluorescent tubes (control) during all days of hospitalization. Subjects will be blinded to the intervention as they will not be told if they are admitted to an intervention patient room or a control patient room. The study includes subjects who are diagnosed with dementia (mild-moderate) or older adults (+65 years) who have cognitive impairments. Subjects will be screened before inclusion using a mini-mental state examination (MMSE) (3) and clinical examination by trained specialists.

To test the effect of exposure to circadian light, the study is designed as a single-centre exploratory parallel-arm randomized controlled trial. The trial will be thoroughly reported according to the CONSORT (4) statement extended guidelines

The primary outcome measure is cortisol levels measured in saliva samples. The samples are collected two times each day during hospitalization. One sample is collected at <30 minutes after the subject has woken in the morning (during the morning cortisol peak) and one sample in the evening when the highest level is expected.

Secondary outcome measures are delirium rates, length of admission, need for escape prevention, mortality, use of antipsychotics and adverse advent e.g. patient related fall incidents. Delirium rates are collected by performing a confusion assessment method (CAM) (5) score two times a day. Additional measurements are collected in patient charts.

To reach sufficient power, we estimate that 80 subjects will be included in the study. 40 subjects are admitted to the intervention room (with naturalistic lighting), and 40 are admitted to the control room (traditional/standard lighting conditions).

Inclusion Criteria:

• Patients with a recognized dementia diagnosis by the time of admission
• Patients who, during admission, are found to have cognitive impairments
• Patients who, during admission, are found to have delirium

Exclusion Criteria:

• Patients with psychiatric conditions which in and of themselves might account for the patients' cognitive impairment will be excluded
• Inability to speak (aphasia)
• Patients with a linguistic or cultural background other than Danish
• Patients with ongoing abuse of alcohol, narcotics or sedative pharmaceutics
• Patients with impaired level of consciousness due to other causes than delirium.