Priming for depression in a dimly lit world

St. Hedwig Hospital and Charité–Universitätsmedizin Berlin researchers report that repeated mornings spent under dim indoor light in healthy young adults raised afternoon and evening cortisol and reshaped sleep in ways known from depressive illnesses.

Urban life in dim light

Depressive disorders are often linked with hyperactivity of the hypothalamic–pituitary–adrenal axis, with cortisol levels that stay elevated into the afternoon and early evening instead of reaching their lowest levels, typical of early evening.

Sleep in depressive illness often carries its own fingerprint. Changes in REM sleep and a shift of slow wave sleep from the beginning of the night toward later phases have been described as biological markers of depression.

Many urban residents spend about half of their daytime below 25 lux in a “Living in Biological Darkness,” with evidence that low daytime light exposure coincides with shorter REM latency and shorter late REM periods.

In the study, “Living in biological darkness III: Effects of low-level pre-midday lighting on markers of depression in healthy subjects,” published in the Journal of Psychiatric Research, researchers used a randomized morning light intervention to test how low daytime lighting shapes cortisol levels and sleep in healthy volunteers.

A total of 20 adults joined the experiment in early spring, with a mean age of 24 ± 3 years, an age range of 19–30 years and an even split by sex. Participants kept habitual bedtimes while accelerometers tracked their sleep–wake schedules. Cortisol markers were assessed in regular urine samples and in saliva over 24 hours.

Morning light as intervention

Morning hours formed the core of the intervention. One group spent six mornings from 8:00 to 12:00 under low intensity incandescent light at 2,700 K, 55 lux and 12 melanopic lux. A second group spent the same hours under higher intensity fluorescent light at 3,500 K, 800 lux and 481 melanopic lux.

Overnight sleep was recorded using polysomnography on four nights. Recordings used central and occipital EEG derivations, and slow wave activity was calculated for six hours from sleep onset across the first three normalized NREM–REM cycles.

Hormone rhythms under dim light

Evening cortisol between 19:00 and 23:00 started at nearly identical levels in both groups. After several dim incandescent mornings, levels rose sharply in the low-light group. Values climbed significantly relative to baseline and stood significantly higher than in the brighter-light group, with very large effect sizes.

In the rest of the waking period, cortisol markers changed little between lighting conditions. Morning and daytime levels outside the evening window remained similar in both groups, and apparent differences did not reach statistical significance.

In the 24-hour saliva series, afternoon cortisol at 16:00 rose under dim light compared with brighter light, while samples at noon and the following morning again showed no significant separation between groups.

Melatonin followed the expected day–night rhythm in each group, without any detectable shift in timing under either lighting condition.

Sleep during nights after dim light

Nights after dim incandescent mornings grew shorter. Total sleep time dropped by roughly 25 minutes in that group, while sleep in the brighter fluorescent group stayed close to baseline duration. Sleep efficiency, sleep onset, REM onset and the broad mix of sleep stages showed no meaningful differences between lighting conditions.

Slow waves and REM timing

Patterns inside the night shifted as well. Under dim morning light, deep slow wave activity moved away from the early part of sleep and into a later NREM–REM cycle. Less slow wave activity appeared in the second cycle and more in the third, with large differences when compared to the brighter light conditions.

Brighter fluorescent mornings brought a different change. REM sleep expanded, especially toward the end of the sleep period, and reached higher amounts than both that group’s own baseline and the dim light group. REM density and REM latency stayed stable across conditions.

Daytime sleepiness and mood

Vigilance on the reaction time task stayed steady. Morning and evening performance looked similar after dim incandescent mornings and after brighter fluorescent mornings.

Subjective ratings told a different story. By day eight, participants who had spent their mornings in low light reported more sleepiness and more sadness than those under brighter light, on the same 10 cm scales that had shown little separation earlier in the week. Evening ratings also drifted apart. Sleepiness climbed more steeply across the evening in the brighter light group, while the dim light group followed a flatter course.

Signals of depression vulnerability

Patterns across hormones and sleep pointed toward a shared direction. Afternoon and evening cortisol ran higher after repeated dim mornings. Nights grew shorter in that group, and deep slow wave activity shifted away from the second NREM–REM cycle toward the third. Brighter mornings, in contrast, brought longer REM sleep late in the night without changes in REM latency or density.

Insomnia often appears before depressive episodes, and an insomnia-like pattern emerged here under low morning light with reduced total sleep time, less restorative sleep earlier in the night and more time in lighter stages. Raised cortisol later in the day and a late shift in slow wave activity match changes frequently described in depressive illnesses.

Rethinking everyday lighting

Office lighting, classroom fixtures and home lamps often create conditions similar to the low incandescent mornings in this protocol. Considering that long stretches of time over a life can unfold with light levels that fall far below natural daylight, the findings point toward a practical response to “Living in Biological Darkness.”

Bringing natural daylight or bright artificial light into schools, workplaces and residential settings, including nursing homes, could strengthen circadian signals and reduce vulnerability linked to dim environments.

© 2025 Science X Network