Don’t leave a light on for me: commentary on Kim et al. “Light at night in older age is associated with obesity, diabetes, and hypertension”

In the 20th century, widespread electrification and relatively inexpensive electric lighting transformed the daily pattern of life for millions of people across the globe. Humans were suddenly able to manipulate their light-dark cycle, no longer consigned to remain in the dark throughout each night. The benefits of access to round-the-clock inexpensive lighting have long been clear, but the negative impacts of light at night (LAN) on human health have only recently been studied.

Although we still are influenced by the solar cycles [1], there is extensive evidence that the timing of sleep has shifted later, resulting in more light exposure in the evening hours between sunset and bedtime [2, 3]. In addition, many individuals experience additional exposure to LAN due to working at night, sleep disruptions, and using light-emitting electronic devices before sleep and during nighttime awakenings. In addition to the impact of LAN while we are awake, there is accumulating evidence that exposure to light even while sleeping can have negative impacts on health. Such exposure is also widespread, both from light sources inside the home and those from outside, particularly in urban areas. Both cross-sectional and prospective studies have found that more LAN is associated with a greater prevalence of obesity and weight gain, and even with the risk of coronary heart disease, but most of those prior studies have focused on the impact of LAN on children, adolescents, and young to middle-aged adults. Whether older adults are exposed to similar amounts of LAN while sleeping, and whether there are negative health outcomes from that exposure is not known.

In this issue of SLEEP [4], Kim et al. explored the association between exposure to LAN and cardiovascular disease (CVD) risk in a population of 552 community-dwelling older adults. Using 1 week of activity and light data collected from a wrist-worn device by participants in the Chicago Healthy Aging Study (CHAS), they categorized participants into LAN and No-LAN groups based on whether they had light exposure during their least active 5-hr episode each day. The LAN group had later bedtimes, shorter sleep, more wake after sleep onset, and lower sleep efficiency than the No-LAN group. They then tested for associations between LAN exposure and health data collected from the same participants using standardized questionnaires, revealing a significant association between LAN and increased prevalence of obesity, diabetes, and hypertension, even after adjusting for demographic characteristics, sleep, and other activity variables. Upon further dividing the LAN group into Low- and High-LAN subgroups, the exploratory analysis found no significant dose-dependent effect of LAN; there were no significant differences between the LAN subgroups in CVD risk factors, although both remained more likely than the No-LAN group to have obesity, diabetes, and hypertension. No association was found between CVD risk and other measures of 24-hr light exposure patterns such as mean/median light intensity, peak intensity, and LAN timing.

The findings by Kim et al. join a growing body of evidence linking exposure to LAN with a wide range of negative health outcomes, including cancer [5], depression [6], and cardiometabolic diseases such as diabetes and overweight/obesity [7]. Previous cross-sectional studies conducted in Nigeria [8] and South Korea [9] have reported an increased prevalence of overweight and obesity in adults living in regions with high outdoor LAN as measured by satellite imagery. Similarly, large studies in adult women in the United Kingdom [10] and the United States [11] have found significant positive associations between self-reported bedroom light levels and higher prevalence of obese or overweight body mass index. In Japan, cross-sectional studies in a population of adults with bipolar disorder [12] and in community-dwelling older adults [13] have also uncovered associations between objectively assessed bedroom LAN and obesity, as well as other measures of CVD risk.

These new data reveal that more than half of the older adults in this study were exposed to LAN during the 5 hr of least activity, and they had greater levels of obesity, hypertension, and diabetes, all risk factors for cardiovascular disease. Potential mechanisms by which these effects may occur include circadian disruption, altered food intake patterns, melatonin suppression, and autonomic dysregulation. While the specific mechanism(s) by which LAN leads to these increased CVD risk factors is not known and the direction of causality may be in question, the study highlights the importance of considering LAN in understanding metabolic and cardiovascular health.

Future interventional studies should be designed to explore each of the potential mechanisms by which LAN can lead to metabolic and cardiovascular dysfunction, and those studies should be carried out in older as well as young and middle-aged adults. More observational studies of LAN exposure and health outcomes that use objectively measured individual light exposure, as in the present study, may provide additional insight into whether LAN exposure has a dose-dependent effect. Finally, the present study suggests that reducing LAN is a potentially modifiable risk factor that could improve metabolic and cardiovascular health. Window shades and dark curtains can be used to reduce or eliminate light from outside the home, while turning off light sources inside the bedroom, and/or wearing a sleep mask are relatively easy and inexpensive ways to reduce LAN, and can be tested for efficacy and acceptability.

We would like to thank Kim et al. for this important report, which will promote more widespread recognition of the importance of light exposure at night for metabolic and cardiovascular health, and which we hope will stimulate further research in the area of LAN and human health.

Funding

The authors were supported in part by National Institutes of Health (NIH) grant R01 DK127254. J.F.D. was also supported by NIH grant R01 AG044416.

Disclosure Statement

None declared.