Approved registration and protocol.

<div><p>Background</p><p>The impact of light exposure on mental health is increasingly recognised. Modifying inpatient evening light exposure may be a low-intensity intervention for mental disorders, but few randomised controlled trials (RCTs) exist. We report a large-scale pragmatic effectiveness RCT exploring whether individuals with acute psychiatric illnesses experience additional benefits from admission to an inpatient ward where changes in the evening light exposure are integrated into the therapeutic environment.</p><p>Methods and findings</p><p>From 10/25/2018 to 03/29/2019, and 10/01/2019 to 11/15/2019, all adults (≥18 years of age) admitted for acute inpatient psychiatric care in Trondheim, Norway, were randomly allocated to a ward with a blue-depleted evening light environment or a ward with a standard light environment. Baseline and outcome data for individuals who provided deferred informed consent were used. The primary outcome measure was the mean duration of admission in days per individual. Secondary outcomes were estimated mean differences in key clinical outcomes: Improvement during admission (The Clinical Global Impressions Scale–Improvement, CGI-I) and illness severity at discharge (CGI-S), aggressive behaviour during admission (Broset Violence Checklist, BVC), violent incidents (Staff Observation Aggression Scale-Revised, SOAS-R), side effects and patient satisfaction, probabilities of suicidality, need for supervision due to suicidality, and change from involuntary to voluntary admission. The Intent to Treat sample comprised 476 individuals (mean age 37 (standard deviation (SD) 13.3); 193 (41%) were male, 283 (59%) were female). There were no differences in the mean duration of admission (7.1 days for inpatients exposed to the blue-depleted evening light environment versus 6.7 days for patients exposed to the standard evening light environment; estimated mean difference: 0.4 days (95% confidence interval (CI) [−0.9, 1.9]; <i>p</i> = 0.523). Inpatients exposed to the blue-depleted evening light showed higher improvement during admission (CGI-I difference 0.28 (95% CI [0.02, 0.54]; <i>p</i> = 0.035), Number Needed to Treat for clinically meaningful improvement (NNT): 12); lower illness severity at discharge (CGI-S difference −0.18 (95% CI [−0.34, −0.02]; <i>p</i> = 0.029), NNT for mild severity at discharge: 7); and lower levels of aggressive behaviour (difference in BVC predicted serious events per 100 days: −2.98 (95% CI [−4.98, −0.99]; <i>p</i> = 0.003), NNT: 9). There were no differences in other secondary outcomes. The nature of this study meant it was impossible to blind patients or clinical staff to the lighting condition.</p><p>Conclusions</p><p>Modifying the evening light environment in acute psychiatric hospitals according to chronobiological principles does not change duration of admissions but can have clinically significant benefits without increasing side effects, reducing patient satisfaction or requiring additional clinical staff.</p><p>Trial registration</p><p>Clinicaltrials.gov <a href="https://clinicaltrials.gov/ct2/show/NCT03788993" target="_blank">NCT03788993</a>; 2018 (CRISTIN ID 602154).</p></div>