Abstract: Life on Earth has adapted to a consistent 24-h solar cycle. Circadian rhythms in physiology and behavior remain synchronized to the environment using light as the most potent entraining cue. During the past century, however, the widespread adoption of electric light has led to `round-the-clock’ societies. Instead of aligning with the environment, individuals follow artificial and often erratic light cycles created by social and work schedules. In particular, exposure to artificial light at night (LAN), termed “light pollution”, has become pervasive over the past 100 years. Virtually every individual living in the U.S. and Europe experiences this aberrant light exposure, and moreover about 20% of the population performs shift work. LAN may disrupt physiological timekeeping, leading to dysregulation of internal processes and misalignment between behavior and the environment. Recent evidence suggests that individuals exposed to excessive LAN, such as night shift workers, have increased risk for depressive disorders, but the biological mechanism remains unspecified. In mammals, intrinsically photosensitive retinal ganglion cells (ipRGCs) project light information to (1) the suprachiasmatic nucleus (SCN) in the hypothalamus, regulating circadian rhythms, and (2) to limbic regions, putatively regulating mood. Thus, LAN has the potential to affect both circadian timekeeping and mood. In this dissertation, I present evidence from rodent studies supporting the novel hypothesis that night-time exposure to light disrupts circadian organization and contributes to depressed mood. First, I consider the physiological and behavioral consequences associated with unnatural exposure to LAN. The effects of LAN on circadian output are considered in terms of locomotor activity, the diurnal cortisol rhythm, and diurnal clock protein expression in the brain in Chapter 2. The influence of LAN on behavior and brain plasticity is discussed, with particular focus on depressive-like behavior (Chapter 3) and effects of SSRI treatment (Chapter 4). Effects of LAN on structural plasticity and gene expression in the brain are described, with emphasis on potential correlates of the depressive-like behavior observed under LAN in Chapter 5. Given the prevalence of LAN exposure and its importance, strategies for reversing the effects are offered. Specifically, eliminating LAN quickly reverses behavioral and physiological effects of exposure as described in Chapter 5. In Chapter 6 I report that administration of a pharmacological cytokine inhibitor prevents depressive-like behaviors in LAN, implicating brain inflammation in the behavioral effect. Finally, I demonstrate in Chapter 7 that exposure to red wavelength LAN reduces the effects on brain and behavior, suggesting that LAN acts through specific retinal pathways involving melanopsin. Taken together, these studies demonstrate the consequences of LAN, but also outline potential avenues for prevention or intervention.

Abstract: BACKGROUND: Recent advances in understanding the fundamental links between chronobiology and depressive disorders have enabled exploring novel risk factors for depression in the field of biological rhythms. Increased exposure to light at night (LAN) is common in modern life, and LAN exposure is associated with circadian misalignment. However, whether LAN exposure in home settings is associated with depression remains unclear. METHODS: We measured the intensities of nighttime bedroom light and ambulatory daytime light along with overnight urinary melatonin excretion (UME) in 516 elderly individuals (mean age, 72.8). Depressive symptoms were assessed using the Geriatric Depression Scale. RESULTS: The median nighttime light intensity was 0.8lx (interquartile range, 0.2-3.3). The depressed group (n=101) revealed significantly higher prevalence of LAN exposure (average intensity, >/= 5 lx) compared with that of the nondepressed group (n=415) using a multivariate logistic regression model adjusted for daytime light exposure, insomnia, hypertension, sleep duration, and physical activity [adjusted odds ratio (OR): 1.89; 95% confidence interval (CI), 1.10-3.25; P=0.02]. Consistently, another parameter of LAN exposure (duration of intensity >/= 10 lx, >/= 30 min) was significantly more prevalent in the depressed than in the nondepressed group (adjusted OR: 1.71; 95% CI, 1.01-2.89; P=0.046). In contrast, UME was not significantly associated with depressive symptoms. LIMITATION: Cross-sectional analysis. CONCLUSION: These results suggested that LAN exposure in home settings is significantly associated with depressive symptoms in the general elderly population. The risk of depression may be reduced by keeping nighttime bedroom dark.