LED Roadway Lighting: Impact on Driver Sleep Health and Alertness (2021)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

48 The goal of the project was to determine the impact of LED roadway lighting on driver sleep health and alertness. Five major findings were evident. First, there were no differences between the 4000 K LED roadway lighting, 2100 K HPS roadway lighting, and no roadway lighting in terms of salivary melatonin suppression from 1:00 AM to 3:00 AM. Second, some differences were observed in the detection and color recognition distances of objects across the various roadway lighting conditions. Specifically, for the 2100 K HPS–High condition, an increase in the exposure time resulted in decreasing the detection and color recognition distances. Third, no differences between any of the roadway lighting conditions were observed for PERCLOS or SDLP. Fourth, there were no differences in the KSS scores across any of the roadway lighting conditions. Fifth, illuminance dosages in the roadway lighting conditions were dependent on the light level and were lower than the illuminance dosages from consumer electronic devices for the same 2-hour exposure. Neither the three levels (high, medium, and low) of 4000 K LED roadway lighting nor the 2100 K HPS–High lighting condition affected salivary melatonin suppression, as evidenced by the lack of statistically significant differences in melatonin suppression between any of the electrical roadway lighting conditions and no roadway lighting. Numerous studies have dem- onstrated melatonin suppression as a result of light exposure in laboratory conditions using monochromatic light sources (Brainard et al., 2001; Lockley et al., 2003; Thapan, Arendt, & Skene, 2001) or broadband white sources at very high light levels of 1000 lux and higher (Wright et al., 2001). This shows that light exposure from any of the roadway lighting conditions, at the highest level of 1.9 lux from 4000 K LED–High, is not strong enough to elicit a detectable salivary melatonin suppression for drivers for the 2-hour exposure from 1:00 AM to 3:00 AM. Statistically significant differences were only observed between the positive salivary control study performed in the laboratory and all of the roadway exposure conditions. The melatonin suppression observed in the positive laboratory control study, which employed a bright white light (1000 µW/cm2; 3500 lux), is consistent with existing research that bright white light can suppress plasma and salivary melatonin in humans (Brainard et al., 2015). The IES recommends the highest street lighting of 1.2 cd/m2 (Illuminating Engineering Society of North America, 2018). Two of the tested street lighting conditions exceeded this luminance by 25% (1.5 cd/m2). As shown in Table 12, the highest average measured subject illuminance during exposure to the four electrical street lighting conditions ranged from 1.1 to 1.9 lux. The average measured participant corneal illuminance during exposure to driving conditions with no electrical street lighting was 1.4 lux. The current study demonstrates that the relatively dim exposures from all the tested roadway lighting environments for drivers do not suppress salivary melatonin from 1:00 AM to 3:00 AM. There is a well-known dose-response relationship between the intensity of light and the resultant magnitude of melatonin suppression (Brainard et al., 1988; Zeitzer et al., 2000). While C H A P T E R   5 Discussion

Discussion 49   this study was not designed to test for the full dose-response of light on melatonin suppression for these polychromatic light sources, certain comparisons can be drawn to existing dose- response curves on polychromatic light sources (Brainard et al., 2015). In these dose-response curves, there is a subthreshold level where incident light causes little to no melatonin sup- pression; a threshold response where increasing light levels cause ever-increasing suppression of melatonin; and a saturation level at which increasing the light level can no longer cause further suppression of melatonin—the response is maximized. In the context of the data pre- sented here, the salivary melatonin data from the positive laboratory control (3500 lux) can be characterized as a saturation melatonin suppression response. In contrast, all of the resultant roadway salivary melatonin levels can be characterized as subthreshold melatonin suppression responses. Differences in the visual performance measures across the roadway lighting conditions were dependent on the duration of exposure as well the color of the object being detected, as evi- denced by the statistically significant differences in detection and color recognition distance. Under the 2100 K HPS–High lighting condition, both detection and color recognition dis- tances showed a decrease with an increase in the exposure time. Such decreases in the visual performance with an increase in exposure time were not observed across any of the 4000 K LED or the no-roadway-lighting conditions. The decrease in the detection and color recognition distance under the 2100 K HPS–High condition could be attributed to fatigue as a result of the relative non-uniformity of the 2100 K HPS—uniformity ratio (UR) = 7.5—over the 4000 K LED (UR = 4.0). A non-uniform roadway has a higher number of darker and lighter bands than a more uniform roadway, and longer exposure to these alternating darker and brighter bands can fatigue the driver more than a uniform roadway. More research is required to understand the reasons for the visual performance decrease in the 2100 K HPS with an increase in exposure time. Differences in the visual performance of the lighting conditions were also observed for the red and yellow targets. The 2100 K HPS–High lighting condition had longer detection and color recognition distances for the red targets; this could be attributed to the higher spectral reflectance (ρHPS = 0.39 versus ρLED = 0.22) of the red targets in the 2100 K HPS–High roadway lighting condition than in any of the 4000 K LED roadway lighting conditions. For the yellow targets, the differences in detection and color recognition distances were primarily dependent on light levels rather than on the light source spectral power distribution, as evidenced by the lack of significant statistical differences between the 2100 K HPS–High and the 4000 K LED– High conditions. Major statistical differences between the detection and color recognition distances were observed between the highest light levels (2100 K HPS and 4000 K LED) and no-roadway-lighting condition, and between no roadway lighting and 4000 K LED–Medium and –Low light levels. PERCLOS and SDLP were not affected by any roadway lighting conditions, as evidenced by the lack of statistical significance in the respective LMMs. For PERCLOS, on average the participants’ eyes were closed approximately 30% of the time across all the roadway lighting conditions and times. This shows that in all the roadway lighting conditions, including no-roadway-lighting, participants exceeded the 12% eye closure threshold that is associated with moderate or greater drowsiness based on existing research (Dasgupta et al., 2013; Hanowski et al., 2008; Wierwille et al., 1994). This result shows that under all exposure conditions participants were equally drowsy. For SDLP, participants in the 2100  K HPS–High condition exhibited numerically higher, albeit not statistically significant, SDLP (more drowsy) in the first 2 time periods than any of the other 4000 K LED or no-roadway-lighting conditions. The SDLP in the last 2 periods were numerically lower (but not statistically significant) in the no-roadway-lighting condition than in any of the roadway lighting conditions. These results indicate an interesting trend in the

50 LED Roadway Lighting: Impact on Driver Sleep Health and Alertness SDLP because of the presence of roadway lighting. Future research in the area with larger sample sizes could result in a better understanding of this phenomenon. KSS scores were significantly affected by the light condition and time. During the positive control session (1000 µW/cm2; 3500 lux), the light exposure was brightest and it resulted in the lowest mean KSS score. This result is consistent with numerous literature reports of light at night enhancing subjective alertness (Hanifin et al., 2019; Lockley et al., 2006). There were no clear trends, however, on the effects of dimmer roadway lighting conditions on KSS scores. Specifically, there were no statistical differences in the KSS scores of any of the road light- ing conditions (1.1–1.9 lux) and no-roadway-lighting conditions (0.8 lux). Statistical differ- ences were only observed between the following pairs: 4000 K LED–High and positive control, 4000 K LED–Medium and positive control, and no-roadway-lighting and positive control. KSS scores in the road conditions were higher than in positive control, indicating that the drivers were sleepier in the road exposure conditions, irrespective of the light level, than in the positive laboratory control. It should be noted, however, that the mean KSS score (M = 6.3, SD = 1.9) in the positive control was still greater than 6, which is the anchor for “some signs of sleepi- ness” on the KSS scale. The effect of time on KSS scores was more consistent, as an increase in exposure time was associated with an increase in the KSS score for all exposure conditions. These results show that neither the spectral power distribution of the light source nor the light level in the roadway lighting conditions significantly affected subjective self-report measures of driver sleepiness. Finally, the illuminance dosages from a continuous 2-hour roadway lighting exposure were considerably lower than the illuminance dosages a person experienced in 24 hours or from a 2-hour exposure from most of the consumer electronic devices. The illuminance dosage from 2-hour exposure to a 4000 K LED–High roadway lighting condition is approximately 0.1% of the total illuminance dosage experienced by daytime office workers (as measured in this study). The highest illuminance dosage from a 2-hour exposure to 4000  K LED–High condition (13162.1 lux-s or 1.9 lux/s) was lower than for all consumer electronic devices except the iPad Pro (3456 lux-s or 0.5 lux/s) and Kindle Paperwhite (12456 lux-s or 1.5 lux/s) in the dark mode for the same 2-hour duration. This also assumes that a person is continually driving in a lighted section for a 2-hour period, which is highly unlikely unless drivers of public transit in major cities are taken into account. In contrast, a 2-hour exposure from a consumer electronic device is a very common occurrence today. Further, to get the same illuminance dosage from a 4000 K LED–High roadway lighting condition as from a 2-hour exposure to a computer monitor and a television in dark mode, a person would have to be in the roadway lighting condition for 17.8 and 5.7 hours, respectively. The potential for melatonin suppression from consumer elec- tronic devices such as television, monitors, and tablets is considerably higher than LED roadway lighting and is supported by published research showing that light exposure from e-readers at 31.73 lux (Chang et al., 2015) and LED computer monitors at less than 100 lux (Cajochen et al., 2011) significantly suppressed melatonin. Overall, the results from the current study show that LED roadway lighting does not signifi- cantly affect driver salivary melatonin and subjective and objective alertness from 1:00 AM to 3:00 AM. Although the results indicated a decrease in the visual performance for the 2100 K HPS light source, those results were not confirmed by other objective (PERCLOS and SDLP) or subjective measures (KSS scores) of alertness. The drop in the visual performance of 2100 K HPS could also be attributed to its high uniformity ratio (more non-uniform; URHPS = 7.5 versus URLED = 4.0). Major differences in melatonin suppression and subjective alertness were observed between the positive laboratory control and all the roadway lighting conditions. These results show that illuminance dosages required to affect salivary melatonin and subjec- tive and objective alertness are much higher than those experienced in typical roadway lighted environments.

Discussion 51   The application of roadway lighting should be considered in terms of several aspects that affect road users and the built environment. These aspects include road safety, energy conser- vation, road user acceptance, impact on user health, sky glow and light trespass, and impact on the surrounding ecology. This project considered primarily the impact on user health, roadway safety (using visual performance as a surrogate), and alertness. The results show that the spectral content and intensity of roadway lighting do not affect the salivary melatonin suppression or alertness in human participants. As a result, spectral power distribution of the light source is not likely a consideration on the health of roadway users (particularly drivers) as the light dosage in the roadway is too low to elicit a measurable effect. However, other consider- ations should be made concerning the spectrum and intensity of roadway lighting, particularly the effects on flora and fauna (Longcore et al., 2018; Palmer et al., 2018; Palmer et al., 2017). These spectrally based considerations should be included in any recommendations. The use of adaptive lighting techniques in which the lighting system is dimmed in reaction to the needs of roadway users provides additional energy savings and further reduces the negative impact of the spectrum. This study has some limitations. First, only one aspect of sleep health—melatonin suppression—was measured in this study. In the literature, sleep health is measured in a multitude of dimensions, such as the number of awakenings, sleep duration, sleep efficiency, and quality of sleep. To better understand the effect of roadway lighting on sleep health, future studies should explore the impact of roadway lighting exposures on other measures of sleep health. Second, the results of this study only account for acute salivary melatonin suppression from exposure to roadway lighting and no-roadway-lighting conditions from 1:00 AM to 3:00 AM. Studies on other evening nighttime periods would be reasonable to conduct. Further, the long- term effects of daily melatonin suppression are unknown and are beyond the scope of the study. Future research in this area should focus on the health impacts of chronic melatonin suppres- sion as a result of daily exposure to roadway lighting. This is especially important for individuals such as drivers of public transit buses in major cities, who are exposed to roadway lighting during their shifts. Other populations that should be studied are roadway pedestrians and resi- dents living near roadways. Third, in this study only salivary melatonin levels were measured in the roadway exposures because of the safety issues to human subjects in collecting blood in a poorly illuminated, non-sterile, and risky environment such as a moving vehicle. Research has shown that blood sampling is more sensitive than saliva for quantitative evaluation of melatonin. Figure 12 illus- trates the great difference in plasma versus salivary levels of melatonin. Future studies could replicate the roadway lighting exposures in safer and sterile laboratory conditions, enabling the examination of plasma melatonin suppression. This approach would contribute significantly to understanding the effect of roadway lighting exposures on melatonin suppression and other parameters of sleep health. Fourth, in the current study, participants continuously drove in the lighted section of the road for 2 hours. In reality, drivers rarely drive for extended periods in lighted roadway environments. The results from this study are only applicable to drivers who are exposed to the roadway light for 2-hour periods. Understanding melatonin suppression and objective and subjective measures of alertness from intermittent exposures to roadway lighting could be the focus of future studies. Furthermore, the lack of statistical significance observed in mea- sures of alertness (PERCLOS, SDLP, and KSS) could also be attributed to the relatively small sample size (n = 10). Increasing the sample size could result in an increase in power, thereby showing statistically significant differences between light source types.

52 LED Roadway Lighting: Impact on Driver Sleep Health and Alertness Fifth, there was no traffic and no extraneous sources of lighting, such as light from parking lots of commercial establishments, on the Smart Road. These simplifications were used to isolate the effects of roadway lighting on melatonin suppression and alertness. The presence of traffic and extraneous sources of light could act as potential confounding effects by introducing additional sources of light and glare. Thus, these results apply to areas with no extraneous light sources other than roadway lighting. Future studies should evaluate melatonin suppression and alertness in more realistic environments. Further, the roadway illuminance and luminance measurements collected as part of this research effort could also be used to understand drivers’ visual adaptation under different roadway function classes.