National Academies Press: OpenBook

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

Chapter: Chapter 2 - Literature Review

« Previous: Chapter 1 - Introduction
Page 4
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 4
Page 5
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 5
Page 6
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 6
Page 7
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 7
Page 8
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 8
Page 9
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 9
Page 10
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 10
Page 11
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 11
Page 12
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 12
Page 13
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 13
Page 14
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 14
Page 15
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 15
Page 16
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 16
Page 17
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 17
Page 18
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 18
Page 19
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 19
Page 20
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 20
Page 21
Suggested Citation:"Chapter 2 - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2021. LED Roadway Lighting: Impact on Driver Sleep Health and Alertness. Washington, DC: The National Academies Press. doi: 10.17226/26097.
×
Page 21

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.

4 Literature Review As the first task of this project, a literature review was prepared on the current state of knowledge regarding the effects of light on human sleep health and alertness. Through this literature review, gaps in research and the state of knowledge on the effects of light on human health and alertness have been identified. Human Vision Vision is essential for driving. The human visual system consists of two main parts: the optical component (the eye) and the image-processing component (the optic nerve to the visual cortex in the brain). The spectrum of electromagnetic waves from 400 to 700 nano- meters (nm) is called the visible spectrum of light because human eyes are sensitive to those wavelengths. Light reflected from objects enters the eye, then goes through the cornea (the protective cover over the eye), the pupil (the opening that controls the amount of light enter- ing the eye), and the lens, which focuses the light onto a layer of photosensitive cells on the retina. The retina then converts this light into an electrical signal and sends it through the optic nerve to the visual cortex in the brain for processing. The lens is responsible for focus- ing objects at different distances from the eyes; the ciliary muscles in the eye help flatten or round the lens to assist in focusing. The retina consists of a photosensitive layer of cells or receptors, which are sensitive to wave- lengths of electromagnetic radiation from 400 to 700 nm. There are four types of photosensitive receptors on the retina; these can be classified into two major categories called cones and rods. Each photoreceptor has a different kind of photopigment, which makes photoreceptors sensi- tive to four kinds of spectra. The rods have the same pigment; therefore, all rods have the same spectral sensitivity. Cones, on the other hand, have three different kinds of photopigments and consequently have three kinds of spectral sensitivities. The three cone receptors are referred to as short (S), medium (M), and long (L), depending on the region in which they have the greatest sensitivity. The response of the S-cone is centered at 437 nm, which coincides with blue light. Light that activates the S-cone is perceived as blue. The M- and L-cones have a large overlap; their spectral sensitivities peak at 533 and 564 nm. These cones are more responsive to green to yellowish-green light, and a good portion of the tails extend into the region perceived as red (wavelengths > 600 nm). Rods and cones comprise the image-forming receptors in the retina. In addition to the image-forming receptors, the retina contains non-image-forming cells, the ipRGCs. These cells play an important role in regulating circadian rhythms. Maximum circadian sensitivity occurs between 12:00 AM and 3:00 AM. The ipRGCs’ peak sensitivity is around the 480 nm wavelength, which coincides with the blue color wavelengths (Bailes & Lucas, 2013; Brainard et al., 2001; Panda et al., 2005). C H A P T E R   2

Literature Review 5   Photopic, Scotopic, and Mesopic Vision The state of adaptation of the human eye dictates spectral sensitivity, as individual photo- receptors (rods and cones) are active at different luminances. There are three defined states of the human eye’s spectral sensitivity: photopic, scotopic, and mesopic vision. Photopic Vision When the luminance of the human eye’s operating environment is greater than 3 cd/m2, the state of adaptation is termed photopic vision. In this state, cone activity is dominant, resulting in good color vision and fine resolution of detail (Boyce, 2009). Scotopic Vision The human visual system’s state is defined as scotopic when the luminance of the environ- ment is less than 0.001 cd/m2. In this state, only rods are active. Color cannot be perceived, and only shades of gray can be identified (Boyce, 2009). Mesopic Vision This is the intermediate state between photopic and scotopic vision. The human eye is in this state of adaptation when the luminance range is between 0.001 cd/m2 and 3 cd/m2. In this state, both rods and cones are active (Boyce, 2009). As luminance decreases through the mesopic state, cone activity in the fovea decreases, resulting in decreased absolute sensitivity without a major change in spectral sensitivity, until the scotopic state is reached. In the periphery, rods slowly come to dominate with decreasing luminance, resulting in deterioration of color vision and resolution of fine detail. At this point, the spectral sensitivity shifts to shorter wavelengths. Nighttime Driving—Mesopic Vision At night, a driver’s visual system operates under mesopic vision (Bullough & Rea, 2004; Plainis, Murray, & Charman, 2005), as any type of vehicle forward lighting will provide enough light to push the visual system into the mesopic state (Boyce, 2009). Scotopic levels are so low that even the presence of moonlight results in mesopic conditions. A field survey of roadway luminances conducted at various roadway intersections that were not continuously lighted concluded that they were in mesopic vision ranges (He et al., 1997). Human vision is not thoroughly understood in the mesopic luminance range, even though most nighttime driving is performed in this range. A study conducted on mesopic spectral sensitivity revealed that as luminance decreased, the contribution of the rods increased, even for on-axis vision (Várady & Bodrogi, 2006). It is suggested that at mesopic luminance levels, eye fixation is not stable and that eye movements will let the peripheral vision be involved, even when the target is placed on-axis. Another study of visual performance in nighttime driving conditions also revealed that with decreasing light levels, rods play an increasing role, and most visual information in nighttime traffic situations is gained from the periphery (Eloholma et al., 2006). While driving at night, the driver’s visual field changes continuously, resulting in exposure to a wide range of luminance values. This variation results in varying states of the driver’s vision adaptation. A study conducted by Plainis, Murray, and Charman (2005), which examined the state of retinal adaptation under road lighting during nighttime driving, suggested that vision is mediated by the cone pathway at a higher mesopic range (5 lux) and by the rod pathway at

6 LED Roadway Lighting: Impact on Driver Sleep Health and Alertness lower mesopic ranges (0.1 and 0.5 lux). The adaptation rate of the visual system slowed signifi- cantly from the higher to the lower illuminance ranges, resulting in higher reaction times at the lower illuminance ranges than the higher illuminance ranges. There was no significant change in the foveal retinal sensitivity. Importantly, retinal sensitivity and speed of recovery decreased in the periphery, where most of the visual field is used while driving (Owsley & McGwin Jr., 1999). Decreasing retinal sensitivity in the periphery was also observed in a peripheral visual performance study (Gibbons et al., 2015) in static realistic nighttime driving environments, where threshold contrast (contrast at detection) increased for objects farther away from the line of sight (increase in eccentricity or off-axis). However, in dynamic conditions, an increase in eccentricity did not result in statistically significant decreases in detection distances, indicating that the driver’s gaze cannot be controlled in realistic conditions (Gibbons et al., 2015). One might expect the detection distances to decrease, as threshold contrast would be higher to detect the object. Purpose of Roadway Lighting The primary purpose of lighting roadways is to make the surrounding areas more visible to road users (vehicle drivers, pedestrians, bicyclists, and the like). This increased visibility gives road users more information about their surroundings and helps them make better action and reaction judgments. This, in turn, makes the roadway safer and reduces the costs associated with loss of life and accidents. The IES standard for roadway lighting enumerates several benefits to roadway lighting, such as reduced nighttime accidents, increased sense of security, better traffic flows, and increased use of public facilities at night (Illuminating Engineering Society of North America, 2018). There are four main sources of lighting on roadways at night: • Vehicle headlamps, which are compulsory by law. • Fixed overhead lighting, which is regulated by standards from organizations such as IES and the Commission Internationale de l’Eclairage (CIE). • Traffic signal lights and illuminated or retroreflective signs present on roadways. • Light from sources that are not present on the roadway, e.g., light from nearby commercial areas and parking lots. Of these, only fixed overhead lighting sources can be controlled by the roadway lighting designer; the rest are either transient or located and owned by private entities, and are not always guaranteed to be available. Consequently, it is for overhead lighting that roadway lighting designers specify light levels and types. Light Sources for Illuminating Roadways High-intensity discharge (HID) light sources have been widely used for illuminating road- ways in the United States. Examples include two types of sodium vapor lamps: low-pressure sodium (LPS) and high-pressure sodium (HPS). Of these, HPS lamps are more commonly used for roadways because of their higher light output and longer life. HPS lamps’ spectrum is low in red, green, or blue wavelengths, and has a correlated color temperature (CCT, the color associ- ated with a light source) of 2100 K (Murdoch, 2003). Recent advances in using LEDs for road- way lighting have resulted in major reductions in energy consumption, by up to 50%. LEDs have higher luminaire efficacy than conventional HPS lamps. LED luminaires differ significantly from HPS lamps in the way they function, their capabilities, their spectral power distribution, and their luminous efficacy.

Literature Review 7   Light Production in LEDs LEDs are solid-state light sources made from semiconducting materials. There is a wide range of semiconducting materials, which are defined as materials whose electrical conductivity is between that of a metal and an insulator. The electrical conductivity of a semiconducting material can be changed by adding impurities (doping), by introducing an electric field, or by making light incident on the material. An LED is a forward biased, p-n junction doped semiconductor. A p-n junction is an interface between two types of semiconducting materials (p-type and n-type). N-type semiconductors have a higher amount of electron concentrations and are negatively charged. P-type semiconductors have higher hole concentrations and are positively charged. LEDs emit light through a process called injection luminescence. In this process, the electrons and holes travel toward the junction, where they meet each other, fall into a lower energy level, and emit light in the form of a photon. The wavelength, and therefore the color of the light emitted, depends on the band gap of the materials forming the p-n junction. For creating white light, phosphor is added to short-wavelength light (like blue; 450–480 nm) emitted by an LED. LEDs have a longer life span and are more shock resistant compared to HID sources. Because LEDs are point sources, they offer better optical control and can be dimmed easily by controlling the forward current (Murdoch, 2003). Luminous Efficacy in LEDs Luminous efficacy is a measure of how well a light source produces light in the visible spec- trum. It is measured as the ratio of luminous flux to power and is expressed as lumens per watt (lm/W). The luminous efficacies of LEDs are higher than those of HPS light sources. An HPS light source has a luminous efficacy of about 60–130 lumens per watt (Murdoch, 2003), whereas LEDs have luminous efficacies in the range of 37–300 lumens per watt (Cree Inc., 2014; Stouch, 2016). Capabilities of LEDs LEDs have several advantages over HID light sources in their ability to turn on and dim: • LEDs can be easily dimmed to specific light levels (Dyble et al., 2005; Li et al., 2009; Gil- de-Castro, Moreno-Munoz, & de la Rosa, 2013; Jin et al., 2015). • LEDs can be turned on and off rapidly without a need for a warm-up period (Gaston et al., 2012; Wang & Liu, 2007). • LEDs can be fine-tuned for color output during the manufacturing process to achieve a range of spectral output (Liu & Luo, 2011). • Precise cutoffs can be implemented for LED luminaires to increase the control of the light’s focus (Timinger & Ries, 2008). • LEDs have life spans of 50,000 to 100,000 hours and outperform HIDs (Josefowicz, 2012). The dimming and rapid on-and-off benefits of LEDs combined with the capacity to control them remotely have made LED light sources perfect for adaptive roadway lighting. An adaptive roadway lighting system is an “intelligent” control system that dims roadway lighting during periods of low traffic and pedestrian activity. Finely controlling a lighting system this way can be a key method for promoting safety during heightened exposure periods while still lowering the effects of sky glow, glare, and light trespass during other periods. The concept of adaptive lighting has reached roadway lighting after effective application in cities such as Cambridge, MA; Seattle, WA; and San Jose, CA. Adaptive lighting control systems reduce a lighting system’s impact when lighting is less needed and provide considerable energy and maintenance cost savings.

8 LED Roadway Lighting: Impact on Driver Sleep Health and Alertness Adaptive street and roadway lighting systems installed in U.S. cities and towns have resulted in significant energy savings, often from 30% to 46% (Escolar et al., 2014; Lau, Merrett, & White, 2013; Municipal Solid-State Street Lighting Consortium, 2013). In some cases, savings can be much greater. For instance, in California, the city of San Jose and the University of California– Davis converted to LEDs and installed adaptive lighting controls, resulting in between 84% and 87% energy savings (Municipal Solid-State Street Lighting Consortium, 2013). In Massachusetts, the city of Cambridge observed a reduction in total energy of about 80% and saved a half-million dollars per year by converting to LEDs and using an adaptive lighting control system (Echelon, 2016). Continual advancements in lighting and adaptive control technologies will shorten the payback period for these systems and will further increase the amount of energy savings. Spectral Power Distributions of LEDs Because of the differences in methods of producing light, the spectral power distribution of LEDs is different from that of HID light sources. Phosphor-converted LEDs generally have a broad-spectrum output compared to HPS light sources, whose spectrum is low in red (650 nm), green (550 nm), and blue wavelengths (450 nm) (see Figure 1). Correlated color temperature, or CCT, can be achieved via a range of spectral distribu- tion characteristics of a light source and is therefore not always effective for defining a light source’s color (Rea, 2000). Correlated color temperature is measured in kelvins (K) and relates to the color of an absolute blackbody radiator when heated to the associated temperature. In general, CCT is a means of easily describing the color of a light source and does not account for melanopic content or spectral power distribution (Rea & Figueiro, 2016). The CCT of a light source is not a reliable indicator of its spectral content. Figure 1. Spectral power distributions of HPS and LED light sources of different CCTs.

Literature Review 9   Visual Performance Benefits of LEDs Nighttime driving presents many unique hazards that are not issues during the day. Specifically, visibility and contrast sensitivity are greatly reduced. Contrast sensitivity is defined by how well an individual can visually separate an object from its background, an important aspect of nighttime driving in terms of detecting objects or pedestrians in the roadway. There are two main types of contrast: luminance and color. Luminance contrast exists on an achro- matic scale that depends entirely on the amount of light reflecting from the surface of an object or its background; the contrast is the measurable difference between the two. Color contrast is the measurable difference between two colors. In general, nighttime driving on an unlighted road removes color contrast due to the low-light environment, where human vision shifts from a photopic to a mesopic range. When roadway lighting is present, color contrast increases visibility and, consequently, safety. LED light sources have better color-rendering capabilities than HPS light sources and result in better visual performance by increasing color contrast. Visual performance in high-speed driving conditions, such as highway driving, is dominated by foveal vision and relies heavily on cone receptors in the eyes (Gibbons et al., 2012). The additional color contrast rendered by the LED light sources increases drivers’ detection capa- bilities at night (Terry, 2011). During nighttime driving, human eyes are in the mesopic range, where both rods and cones are active. Lighting that enhances color contrast in addition to luminance contrast can increase an object’s visibility on the roadway (Lutkevich, McLean, & Cheung, 2012). Studies that compared the visual performance of LED and HPS sources showed that LED sources provided a benefit to visibility (Clanton and Associates and Virginia Tech Transporta- tion Institute, 2010; Clanton et al., 2014; Gibbons et al., 2010; Mutmansky et al., 2010). These studies were conducted at different test sites in different environments and geometries, so the locations themselves are not directly comparable. In all these studies, a standard gray-colored target of size 18 cm by 18 cm (50% reflectance) was used for detection. The light sources dif- fered based on the manufacturer and experimental location; however, in comparisons made at the same testing sites, LEDs were the best performers in terms of visibility (see Table 1). Direct comparisons to HPS sources indicate that LED light sources, which are typically between 4000 K and 5000 K, significantly outperform HPS and LPS sources in visibility distance. Studies comparing roadway lighting, specifically HPS and LED light sources of differing CCTs, determined that 5000 K luminaires, which consist of shorter wavelengths or bluer light, did not outperform the 4100 K luminaires in terms of color perception or detection distance (Clanton & Associates & VTTI, 2014). These results also showed that CCT is a weak measure for determining a light’s ability to render color or improve color contrast. Figure 2 shows the results of the study indicating the benefit of the 4100 K luminaire for detection distance over other light sources, including those with higher CCTs. The 4100 K luminaire was the best performer for visibility overall. Research has also shown that LED light sources increase peripheral visual performance in drivers at night (Gibbons et al., 2015). An increase in peripheral visual performance helps drivers detect hazards on the off-axis or outside the roadway. A study investigating the effect of spectral power distribution on peripheral object detections found that LED light sources (such as the 3500 K and 6000 K CCTs) performed better than HPS sources when participants were tasked with identifying off-axis objects in a driving scenario (Gibbons et al., 2015). Because of the spectral power distribution of LED light sources, exposure to LED light sources can affect human health and alertness. The following sections of the literature review will address these aspects of lighting.

10 LED Roadway Lighting: Impact on Driver Sleep Health and Alertness Figure 2. Luminaire type and light level by detection distance (wet and dry pavement)—100, 50, and 25—represent the light level as a percentage of the full power. 250 W and 400 W represent the power of the HPS luminaires. 3500 K, 4100 K, and 5000 K represent the correlated color temperature (CCT) of the LED luminaires. ASYM refers to an asymmetrical optic type for an LED luminaire with a CCT of 4100 K (Clanton & Associates & VTTI, 2014). Location Light Source Type CCT ~Avg. Target Detection Distance (ft.) Anchorage, AK LED 4100 K 213 Anchorage, AK LED 4300 K 210 Anchorage, AK Induction 4000 K 174 Anchorage, AK LED 3500 K 167 Anchorage, AK HPS 2000 K 141 San Diego, CA LED 3500 K 135 San Diego, CA Induction 3000 K 131 San Diego, CA HPS 2100 K 128 San Diego, CA Induction 3000 K 125 San Diego, CA LED 3500 K 105 San Jose, CA LED 5000 K 233 San Jose, CA LED 4000 K 223 San Jose, CA Induction 4000 K 197 San Jose, CA HPS 2100 K 193 San Jose, CA LPS 1700 K 190 San Jose, CA LED 3500 K 157 Seattle, WA LED 4100 K 145 Seattle, WA LED 4000 K 138 Seattle, WA LED 5000 K 122 Seattle, WA HPS 2000 K 103 Seattle, WA LED 3500 K 100 Seattle, WA HPS 2000 K 68 Sources: Clanton and Associates and Virginia Tech Transportation Institute, 2010; Clanton et al., 2014; Gibbons et al., 2015; Mutmansky et al., 2010. Table 1. Detection distances of luminaires from evaluations comparing different lighting systems.

Literature Review 11   Circadian Regulation by Light The circadian timing system controls daily rhythms such as sleep and wakefulness, body temperature, hormonal secretion, and other physiological parameters. These inherent rhythms persist even when the organism’s environment remains in a constant state. This indicates that these rhythms are under the control of endogenous oscillators or the circadian timing system. The circadian timing system allows the organism to anticipate and prepare for the profound changes in its natural environment at dawn and dusk. Although light is the primary stimulus for regulating the circadian system (Czeisler, 1999), other external stimuli, such as the timing of sound, temperature, and social cues, may also influence physiological functions (Aschoff, 1981; Wetterberg, 1993). Circadian clocks respond to light at dusk and dawn. The appearance of light at dawn or the diminishment of light at dusk causes changes to the phase of the clock, to keep it synchronized. Light detected in the early morning advances the circadian oscillator, while light of an equal amount in the evening causes a delay. This response to light information given at different times of the day results in a phase response curve, which has been well characterized in humans (Czeisler et al., 1989; Khalsa et al., 2003; Rüger et al., 2012). Also, the existence of a light phase response curve is evident in the vast majority of terrestrial plants and animals studied, but varies greatly in shape between organisms (Johnson, 1990). The suprachiasmatic nuclei (SCN) of the hypothalamus are the primary pacemakers for the human and mammalian circadian systems. The SCN distribute information about light, darkness, and biological time to other major central nervous system control centers. Afferent projections from a multitude of nuclei reach the SCN, while efferent projections from the SCN influence endocrine responses, autonomic activity, and behavioral regulation (Dunlap, Loros, & DeCoursey, 2004; Klein, Moore, & Reppert, 1991). Further, many tissues and organs in the body can generate circadian rhythms independently from the SCN (Yamazaki et al., 2000; Yoo et al., 2004). There is evidence that the rhythms of many of the peripheral oscillators are regulated both to each other and the external environment, and humoral and neural signals are generated by the SCN, which regulate the phases of peripheral oscillators (Kalsbeek, Perreau- Lenz, & Buijs, 2006; Pezuk et al., 2010; Prasai et al., 2011). Circadian, Neuroendocrine, and Neurobehavioral Photoreception Intrinsically photosensitive retinal ganglion cells (ipRGCs) are the initial interaction point for nonvisual responses between the nervous system and the light environment. Studies have confirmed that the SCN receive environmental photic input from a specialized subset of ipRGCs (Berson, 2002; Gooley et al., 2001; Hannibal et al., 2002; Hattar et al., 2002; Provencio et al., 2000). IpRGCs respond to light on their own when separated physically or blocked by drugs from receiving input from other neurons (Berson, 2002; Dacey et al., 2005; Hattar et al., 2002; Lucas et al., 2003). These unique cells comprise 1%–3% of all retinal ganglion cells, and their light responses appear to parallel those for circadian entrainment and melatonin suppression (Berson, 2002; Hattar et al., 2003). It has also been shown, however, that rod and cone photoreceptors still play a role in this physiology. Mice genetically altered to lack melanopsin show that the classical rod and cone photoreceptors can compensate for the loss of melanopsin and at least partially mediate light- induced circadian, neuroendocrine, and neurobehavioral responses (Lucas et al., 2003; Lucas et al., 2014; Panda et al., 2002; World Health Organization, 2010). In humans, cones are believed to contribute substantially to circadian photoentrainment at low intensities and at the initiation

12 LED Roadway Lighting: Impact on Driver Sleep Health and Alertness of light exposure, as described in a series of experiments in which spectral content and irradiance were manipulated (Gooley et al., 2010). Different models of interaction between the melanopsin ganglion cells and classical rod and cone visual photoreceptors have been proposed, including the ideas that these photoreceptors are additive, stimulatory, opponent, or dynamic over time and intensity (Figueiro et al., 2004; Gooley et al., 2010; Revell & Skene, 2007). Projections from the ipRGCs form the origin of the retinohypothalamic tract (RHT). The RHT is the primary nerve pathway to the circadian clock synchronizing circadian rhythms to the light-dark cycles of the environment (Moore, 1983). This pathway acts to convey informa- tion about external light conditions from the retina to several areas of the brain, including the SCN, a primary site of the biological clock or central timing system in the brain (Moore & Lenn, 1972; Moore, Speh, & Card, 1995). Even though it is known that the RHT projects to other areas of the brain (Pickard & Silverman, 1981) and the SCN receives inputs from other visual areas of the brain, the retinal projection to the SCN via the RHT seems to be the primary pathway for light-dark synchronizing in mammals. Pineal Gland and Melatonin Production A nerve pathway carrying light information extends from the SCN to the pineal gland through a rather long, indirect pathway, making several connections in several brain regions (Moore, 1983). By way of this neuroanatomy, cycles of light and dark, which are perceived through the eyes, entrain SCN neural activity, which, in turn, entrains the rhythmic synthesis and secretion of melatonin from the pineal gland. Melatonin is a hormone synthesized and released by the pineal gland during the hours of darkness in all species studied to date. Numerous functions are associated with melatonin, chief among them circadian rhythm regulation and seasonal reproduction (Arendt, 1995). Other physiological roles of melatonin include protection of tissues from oxidation (Reiter et al., 1994) and cancer-suppressing effects (Blask et al., 2002). It is used as a therapeutic agent for jet lag following rapid time-zone changes, shift work sleep disturbances (Arendt et al., 1997; Skene, Deacon, & Arendt, 1996), and metastatic cancer (Lewy, Cutler, & Sack, 1999). Marketed as a dietary supplement, it is widely available in some countries, including the United States (Bonn, 1996). The half-life of melatonin in plasma is about 10 minutes due to rapid clearing in the liver (Iguchi, Kato, & Ibayashi, 1982). Melatonin is not stored systemically anywhere in the body, and as a result, melatonin levels in the blood directly mimic the rates of production and metabolism. The very precise and changeable rhythmic production and degradation of melatonin are a very efficient way for transmitting the message of environmental darkness throughout all tissues and organs of the body. The duration of melatonin secretion at night varies in proportion to the length of the dark phase, and in this way signals the season to the organism. The timing of the melatonin rhythm is considered to be the best physiological marker of the circadian phase, and its normal rise in concentration during an evening dim light exposure, named dim light melatonin onset, is used both clinically to characterize circadian rhythm disorders and experimentally to study circadian physiology (Burgess et al., 2016; Klerman et al., 2002; Lewy et al., 1999; Voultsios, Kennaway, & Dawson, 1997). In virtually all species, including humans, high levels of melatonin are secreted during the night and low levels are secreted during the day (Arendt, 1995). In addition to entraining mela- tonin synthesis by the pineal gland, light can acutely suppress melatonin synthesis. The acute light-induced suppression of nocturnal melatonin synthesis was first observed in rats (Klein &

Literature Review 13   Weller, 1972) and has been used in numerous animal and human studies to help determine the biochemical, physiological, and neural mechanisms of melatonin regulation (Brainard et al., 1988; Brainard, Rollag, & Hanifin, 1997; Lewy et al., 1980). Light-Induced Melatonin Suppression The chief factors that affect the degree of melatonin suppression include corneal light inten- sity and the timing and duration of the light exposure, previous light-exposure history, and wavelengths emitted by the light sources. There is a well-known dose-response relationship between the intensity of light and the resultant magnitude of melatonin suppression (Aoki et al., 1998; Bojkowski et al., 1987; Brainard et al., 1988; Brainard et al., 1983; Lynch et al., 1981; McIntyre et al., 1989; Zeitzer et al., 2000). As mentioned earlier, light is the most important cue for entraining the circadian timing system. Light phase response curves have been constructed for a short, 1-hour pulse of bright white light (St Hilaire et al., 2012) as well as for narrow-band, blue LED light (Revell, Molina, & Eastman, 2012). Several human studies showed that a higher-intensity light history over a given period dampens the magnitude of subsequent melatonin suppression by acute light exposure (Hébert et al., 2002; Jasser et al., 2006; Smith, Schoen, & Czeisler, 2004). In one study, 1 week of con- trolled dim light (less than 200 lux) resulted in a 53% subsequent suppression of melatonin by 500 lux white light compared with only 41% after a week of bright light (5000–7000 lux) expo- sure (Hébert et al., 2002). In another study, a light history of 0.5 lux for over 63 hours prior to a 6.5-hour 200 lux white light exposure resulted in a mean melatonin suppression of 85.7%, compared with 71.2% after a constant light history of 200 lux over an equal period (Smith et al., 2004). Several years after Lewy and colleagues discovered that light at 2500 lux can suppress mela- tonin in humans (Lewy et al., 1980), Brainard worked with Lewy and others to determine more precisely the dosages of light needed to suppress melatonin in normal volunteers (Brainard et al., 1988). In that study, six healthy males were exposed to carefully controlled intensities of monochromatic light at 509 nm for 1 hour during the night. The data demonstrated that light at 509 nm affects melatonin in a fluence-response (dose-response) fashion (i.e., the brighter the photic stimulus, the greater the suppression of melatonin). The rigidly controlled, monochro- matic 509-nm exposures, 1.6 µW/cm2 (19 lux), elicited a 37% mean melatonin suppression. Such experimental exposures are not typically encountered in routine living conditions. Multiple laboratories performed wavelength comparison studies on melatonin suppression, using monochromatic (less than 15 nm half-peak bandwidth) or narrow-band (15 nm or greater half-peak bandwidth) light sources. Studies using monochromatic light exposures in humans have shown that melatonin suppression is maximally sensitive to blue light (between 459 and 484 nm) (Brainard et al., 2001; Thapan, Arendt, & Skene, 2001). Exposure to 6.5 hours of mono- chromatic light at 460 nm caused twice the amount of melatonin suppression as 555 nm light (Lockley, Brainard, & Czeisler, 2003). Increasing irradiances of narrow-band blue LED light (λmax 469 nm, half-peak bandwidth 26 nm) can elicit increasing plasma melatonin suppression in healthy subjects (p < 0.0001) (West et al., 2011). In that study, 20 µW/cm2 (19 lux) exposure elicited a statistically significant suppression of melatonin. Further, some work has been done comparing broadband polychromatic light sources, which are described in terms of CCT. Specifically, higher color temperature fluorescent lamps evoked stronger melatonin suppression than lower correlated color temperature bulbs (Morita & Tokura, 1998; Sato, Sakaguchi, & Morita, 2005). More-refined studies comparing poly chromatic light with more short-wavelength energy (8000 K CCT) to typical fluorescent

14 LED Roadway Lighting: Impact on Driver Sleep Health and Alertness lighting (4100 K CCT) for melatonin suppression revealed differences, but only at one of several low-light levels that were compared (Figueiro, Rea, & Bullough, 2006). One study compared 2000-lux light emitted by broad-spectrum white LEDs, blue/green LEDs, and white fluorescent lamps (Wright, Lack, & Partridge, 2001). The results showed no significant differences between light from white LEDs and white fluorescent lamps for acute melatonin suppression, but broad- spectrum blue/green light appeared to have a stronger melatonin suppression effect than both white light sources. Three polychromatic fluorescent light sources differing in relative emis- sion of lighting in the 400–500 nm range showed that increasing irradiance evoked increasing levels of nocturnal melatonin suppression. Comparison of the resulting dose-response curves revealed that polychromatic fluorescent light is more potent for melatonin suppression when enriched in the short-wavelength spectrum (Brainard et al., 2015). Comparisons of LED computer screen outputs show similarly enhanced melatonin suppres- sion with blue-enriched exposures (Cajochen et al., 2011), and lights designed to reduce the short-wavelength output in the melanopic range also show reduced melatonin suppression (Rahman, St Hilaire, & Lockley, 2017). A recent study compared typical domestic and workplace lamps over a short-term exposure, such as one might experience during an extended bathroom visit prior to bed, for effects on melatonin and subjective measures of alertness. Data revealed that the evening rise in salivary melatonin was reduced by lamps containing an abundance of blue wavelengths (Wahnschaffe et al., 2013). Effects of Light on Sleep To obtain the best sleep duration and quality, the timing of the sleep must be appropriately aligned with the timing of the circadian clock. In humans, exposure to light in the evening and early part of the night suppresses the release of the hormone melatonin, which is known to promote sleep (Brainard et al., 1988; Zeitzer et al., 2000). Light exposure also shifts the circadian clock to a later time (Khalsa et al., 2003). Melatonin suppression and phase delay of the clock make it more difficult to fall asleep at night. One study examined the consequences of evening electrical lighting on the human circadian clock (Wright Jr. et al., 2013). It was shown that electrical lighting in the built environment is associated with delayed timing of the circadian clock as compared to a summer natural light-dark cycle experienced during camping. The cir- cadian clock synchronized to solar time such that the beginning of the internal biological night occurred at sunset. These findings emphasize how modern light-exposure patterns contribute to late sleep schedules and disrupt sleep and circadian clocks. Such disruption has clear impact on human health. Direct effects of nighttime lighting on sleep health have been examined both in the laboratory and epidemiologically. For example, Chang and colleagues compared the effects of reading an electronic book on a light-emitting device (LE-eBook) with reading a printed book in the hours before bedtime. Participants reading an LE-eBook took longer to fall asleep and had reduced evening sleepiness, reduced melatonin secretion, and later timing of their circadian clock than when reading a printed book (Chang et al., 2015). Similarly, individuals who slept with a night light of approximately 40 lux had shallower sleep and increased arousals, as well as decreased brain oscillations during sleep (Cho et al., 2013). Both studies illustrate how light exposure in the evening or nighttime affects sleep quality and health. Indoor nighttime light exposure is not the only light source contributing to detrimental sleep health. Two large epidemiological studies reported that light trespass from outside the home was sufficient to disrupt sleep. In a study of more than 19,000 individuals, those living in areas with more outdoor night lighting self-reported less quality and quantity of sleep, as well as more

Literature Review 15   daytime sleepiness (Ohayon & Milesi, 2016). In another cohort of almost 9,000 individuals, symptoms of insomnia and snoring were associated with areas of high light pollution (Koo, Choi, & Jung, 2013). LED technology is becoming a greater presence in our homes, workplaces, and environ- ment. Given the ever-expanding use of electronic devices in the evening, the spectral compo- sition of computer screens that used LED backlight was compared to the spectral composition of computer screens that used a cold-cathode fluorescent lamp (CCFL). The photon flux of blue light of the LED-backlit computer screen was three times higher than that of the CCFL computer screen. A 5-hour evening exposure resulted in a significant suppression in the evening rise of salivary melatonin levels as well as electroencephalogram (EEG) measures of objective sleepiness (Cajochen et al., 2011). These results demonstrate that evening ambient artificial light exposure and the use of devices containing LEDs can delay the circadian clock and acutely sup- press melatonin. Taken together, the laboratory and epidemiological studies add to a growing literature on how nighttime lighting has important implications for sleep health. Direct Alerting Effects of Light at Night Broadband, white fluorescent light sources have been assessed in comparison studies on alertness, neurobehavioral testing, melatonin suppression, and circadian phase changes. Early studies have shown that bright white polychromatic light exposure has been shown to acutely increase alertness and reduce fatigue during light exposure (Badia et  al., 1991; Dawson & Campbell, 1991; French, Hannon, & Brainard, 1990; Myers & Badia, 1993). Bright, overhead, white polychromatic light (Philips FB40/CW/3 cool white fluorescent) at 3000 lux, when com- pared to 100-lux dim white light, increased measures of subjects’ performance on 7 of 10 neuro- behavioral tests during a 16-hour overnight light exposure in nine male subjects (French, Hannon, & Brainard, 1990). A similar study simulating night-shift work compared a 4-hour period of 6000 lux white light to dim white light at less than 200 lux throughout an 8-hour shift in 13 male and female subjects. The group that received the 4 hours of bright light on the first night of the study exhibited a smaller decline in alertness during the second half of the night shift than the group that received dim light the entire time (Dawson & Campbell, 1991). Another larger study compared 43 male subjects in a between-subjects design in four different lighting conditions. An increase in brain wave frequencies associated with alertness and improvement in performance on the neurobehavioral test battery was found for the bright light exposures (Badia et al., 1991). In all of these studies, the light source was not highly char- acterized, if described at all. Importantly, not all studies showed an effect of bright white light on performance and alertness. For example, in a study with 24 male subjects, 13.5-hour light exposures of white fluorescent light at 300, 1500, or 3000 lux were compared for performance on a neurobehavioral test battery (Dollins et al., 1993). Lighting conditions had no effect on performance on a battery of neurobehavioral tests. Performance on the test battery declined on successive testing days, regardless of light condition. Many investigators continued to perform studies on acute, dose-dependent alerting effects to broadband fluorescent light sources presented to subjects at night. Measures taken include sub- jective sleepiness, psychomotor vigilance reaction times, reduced lapses, reduction of attentional failures indicated by electrooculogram-derived slow rolling eye movements, and suppression of theta-alpha activity in waking EEG (Badia et al. 1991; Burgess et al., 2001; Cajochen et al., 2000; Campbell et al., 1995; Daurat et al., 1993; Dawson & Campbell, 1991; Lavoie et al., 2003; Myers & Badia, 1993; Wright et al., 1997). The majority of early work on the alerting effects of light at night on objective measures was broadly consistent with those for melatonin suppression. Some studies tested selective wavelengths on the alerting effects of light (Cajochen et al., 2005;

16 LED Roadway Lighting: Impact on Driver Sleep Health and Alertness Lockley et al., 2006). In one of those experiments, exposure to 6.5 hours of 460 nm mono- chromatic light at night significantly improved auditory reaction time, lowered lapses of atten- tion, and reduced subjective sleepiness compared to the same photon density exposure as 555 nm light (Lockley et al., 2006). High-alpha activity in the waking EEG was elevated and delta activity lowered in response to 460 nm light, consistent with a more alert state (Lockley et al., 2006). Similarly, exposure for 6 hours to dim narrow-band blue light at night (1 lux, 460 nm) sup- pressed EEG delta and theta activity and improved psychomotor reaction time when compared to a dim white light (0.2 lux) or red light (1 lux, 640 nm) control (Phipps-Nelson et al., 2009). There is a well-established dose-response for the alerting effects of white fluorescent light (4000 K) at night for subjective and objective measures of alertness which saturates in the mid- hundreds of lux following exposure to dim light of less than 10 lux for 6.5 hours (Cajochen et al., 2000). There is no equivalent daytime dose-response curve, although limited dose-dependent effects have been described, as outlined here. There are limited comparisons of different spectral power distributions (SPDs) with regard to the effects of light pattern, light history, and light timing. An impact has been shown on the alerting response of changing the color temperature to levels of light experienced in the home environment. Chellappa and colleagues demonstrated enhanced psychomotor performance and subjective alertness during a 2-hour, 30-lux expo- sure to blue-enriched (6500 K) compact fluorescent white light compared to incandescent red- enriched light (2500 K, 3000 K). The effects of the blue-enriched light continued into sleep even after the lights were turned off (Chellappa et al., 2011; 2013). Blue-enriched white light from a computer screen also improved subjective alertness, a range of performance measures, and EEG-derived objective correlates of alertness compared to an SPD reduced in the 454 nm to 474 nm range (Cajochen et al., 2011). In a randomized, within-participant, inpatient study, it was demonstrated that reading from a light-emitting e-reader compared with a printed book induced biological effects on acute subjective and objective evening alertness (Chang et al., 2015). In a large laboratory trial of night-shift workers, who were studied the night after completing a series of night shifts at their workplace, exposure to blue-enriched (17000 K) fluorescent light was shown to improve subjective alertness but not performance or other objective correlates of alertness, compared to an equal photopic lux exposure (approximately 90 lux) of 4000 K fluorescent light during the circadian nadir, when intrinsic sleepiness is highest (Sletten et al., 2017). Similarly, exposure to evening blue-enriched (6500 K) light also improved alertness and performance (Münch et al., 2016). As in the earlier work with bright white light exposure on alertness and performance, not all studies demonstrated consistent alerting and performance responses (Segal et al., 2016; Rahman, St Hilaire, & Lockley, 2017). Interaction of Alertness and Melatonin Multiple studies have shown that light exposure can positively influence alertness. For night- time light-exposure effects on alertness, it has been hypothesized that the mechanism of action is related to the suppression of melatonin (Cajochen, Kräuchi, & Wirz-Justice, 2003; Cajochen et al., 2005; Chellappa et al., 2011; Lockley et al., 2006; Perrin et al., 2004), while others have shown that there is no relationship between melatonin suppression and alertness (Revell et al., 2010; Rüger et al., 2006). Daytime improvements in alertness with bright white light exposure cannot be explained similarly, as daytime melatonin levels are undetectable. Further, another study suggested an alternate pathway, with different neural processing as well as spectral sensi- tivity for the two light responses (Revell et al., 2006). It was observed in this study that exposure to short-wavelength light of 420 nm had a greater effect on subjectively measured alertness and

Literature Review 17   mood than the 470 nm light exposure. The acute effects of light exposure on alertness during the day are unlikely to be mediated by melatonin. Further, it is unclear how alerting or performance effects of light exposure at night are related to melatonin regulation. As delineated, light is a potent stimulus for regulating human circadian, neuroendocrine, and neurobehavioral responses in humans. Because of light’s broad influence on human physi- ology and behavior, light has been used as a therapeutic device. Since the 1980s, bright white light therapy has proven to be an effective therapeutic intervention for patients with seasonal affective disorder (SAD, or winter depression). Additional clinical applications have been explored, including light treatment of various sleep disorders, non-seasonal depression, bulimia nervosa, menstrual cycle problems, and fatigue problems associated with senile dementia, chemo therapy, and traumatic brain injury. In addition, light therapy has been evaluated in nonclinical applications for healthy individuals who experience problems related to shift work, intercontinental jet travel, and spaceflight (Brainard & Hanifin, 2016; Illuminating Engineering Society of North America, 2008; Journal of Biological Rhythms, 2005; Lam & Tam, 2009; Lucas et al., 2014; Wirz-Justice, Benedetti, & Terman, 2013). In the early decades of developing light therapy for both clinical and nonclinical applications, broad-spectrum white fluorescent or incandescent light sources were used. With the advent of solid-state technology, light therapy devices are now being produced with both narrow and broad bandwidths of light emitted by LEDs. This advance has enabled light therapy equipment for clinical applications to be produced in conventionally sized light panels as well as relatively small, portable, or wearable devices (Brainard & Hanifin, 2016). A fundamental principle of modern medicine is that agents that can heal also have the poten- tial to harm. Disruption of the circadian system by inappropriate light exposure has been linked to various disorders and diseases. As an example, in 1987 it was hypothesized that the increased risk of breast cancer in industrialized countries is due, in part, to increased exposure of the population to electrical light at night (Stevens, 1987). The hypothesis was based on the con- cept that exposure to light at night would result in melatonin suppression that would, in turn, increase breast cancer risk. Since then, numerous epidemiological studies have supported that hypothesis, including observations that women who work night shifts are at higher risk; risk has an inverse association with sleep duration; blind women are at lower risk; and across societies, there is a correlation between the incidence of breast and prostate cancer and nighttime ambient illumination as measured by satellite imagery. Generally, the majority of published epidemio- logical studies support the “light, melatonin, breast cancer hypothesis” (Stevens et al., 2013). The majority of shift work involves routine light exposures during the nighttime that have the potential to suppress nocturnal melatonin secretion, disturb circadian rhythms, and inter- fere with normal, healthy sleep (Stevens et al., 2013). It is not clear if circadian disruption due to inappropriate light exposure alone increases the risk of developing cancer, cardiovascular disease, or metabolic disorders. Disruption of the human circadian system usually involves dis- ruption of sleep and/or the disruption of normal melatonin rhythms (Stevens et al., 2013). Sleep deprivation and nocturnal melatonin suppression each have been implicated in the potential health consequences associated with shift work and light exposure at night (Buxton et al., 2012; Morris et al., 2015; Scheer et al., 2009). In general, epidemiological studies are valuable for identifying associations between disease and potential environmental risk factors. It is impor- tant to note, however, that epidemiological studies do not prove causation. Evidence on the cause of disease is best derived from empirical studies. Empirical evidence from both in vitro and in vivo studies with animal models also supports elements of the “light, melatonin, breast cancer hypothesis” (Blask et al., 2011; Stevens et al., 2013). Importantly, it has been shown that human breast cancer xenografts perfused with nocturnal, physiologically melatonin-rich blood collected from premenopausal female volunteers during the night exhibited suppressed breast

18 LED Roadway Lighting: Impact on Driver Sleep Health and Alertness cancer proliferative and metabolic activity. These results identify a definitive nexus between the exposure of healthy premenopausal female human subjects to light at night and the enhance- ment of human breast cancer growth progression via disruption of the circadian melatonin signal (Blask et al., 2011). Based on the epidemiological and empirical literature on breast cancer, the International Agency for Research on Cancer (IARC), which provides assessments of the cancer risk of poten- tial carcinogens for the World Health Organization (WHO), hosted an international meeting of 24 scientists to assess the carcinogenicity of shift work, painting, and firefighting. Based on “limited evidence in humans for the carcinogenicity of shift-work that involves nightwork” and “sufficient evidence in experimental animals for the carcinogenicity of light during the daily dark period (biological night),” the IARC working group concluded that “shift-work that involves circadian disruption is probably carcinogenic to humans” (Straif et al., 2007). Later, in 2010, the WHO published an extensive monograph on the accumulated evidence on shift work and breast cancer and the conclusion that shift work “that involves circadian disruption is probably carcinogenic to humans” (World Health Organization, 2010). As discussed, shift work nearly always involves exposure of workers to light during the night- time hours. Although the 2010 WHO monograph had a primary focus on risk for breast cancer, there have been emergent epidemiological data identifying an association between prostate cancer risk and shift work (Rao et al., 2015; Stevens et al., 2013). Similarly, preliminary empirical data have shown that human prostate cancer xenografts perfused with nocturnal, physiologically melatonin-rich blood collected from young male volunteers during the night exhibited sup- pressed prostate cancer proliferative and metabolic activity (Dauchy et al., 2011; Hanifin et al., 2016). As with breast cancer, these empirical results suggest that exposure of males to light at night causes human prostate cancer growth progression via disruption of the circadian mela- tonin signal. It is important to note that the data on light exposure at night for prostate tumori- genesis are at a much earlier stage compared to the epidemiological and empirical data related to breast cancer tumorigenesis. Two years after the publication of the complete WHO monograph, the AMA published a policy statement on potential health hazards of light at night. Specifically, the AMA identified a need for “further multidisciplinary research on occupational and environmental exposure to light at night, the risk of cancer, and effects on various chronic diseases” (American Medical Association, 2012). Following that, the AMA published a second policy statement from the asso- ciation’s House of Delegates (HOD) entitled Human and Environmental Effects of Light Emit- ting Diode (LED) Community Lighting (Kraus, 2016). That report reached three conclusions: 1. That our American Medical Association (AMA) support the proper conversion to community-based LED lighting, which reduces energy consumption and decreases the use of fossil fuels. (New HOD Policy) 2. That our AMA encourage minimizing and controlling blue-rich environmental lighting by using the lowest emission of blue light possible to reduce glare. (New HOD Policy) 3. That our AMA encourage the use of 3000 K or lower lighting for outdoor installations such as road- ways. All LED lighting should be properly shielded to minimize glare and detrimental human and environmental effects, and consideration should be given to use the inherent ability of LED lighting to be dimmed for off peak time periods. (New HOD Policy) (American Medical Association, 2016) These recommendations—particularly the last—elicited a range of responses from constit- uents in the lighting community (Illuminating Engineering Society of North America, 2017; National Electrical Manufacturers Association, 2016; DOE, 2016). One concern with the third AMA conclusion was that it used the CCT as a metric. Although this is a reasonable concern, it is important to acknowledge that the CCT metric has been strongly promulgated by the light- ing industry for decades and is still used widely for specifying roadway lighting (Illuminating

Literature Review 19   Engineering Society of North America, 2011). To address this concern, the DOE published a table that compared 21 different light sources used in street and area lighting and selected performance characteristics related to their spectral content. This table is reproduced here as Table 2 and includes measurements of CCT, calculations of the percentage of radiant power contained in “blue wavelengths,” and the corresponding scotopic and melanopic contents rela- tive to an HPS baseline, normalized for equivalent lumen output (DOE, 2016). There is no standardized, single measurement unit available for quantifying light for circa- dian, neuroendocrine, and neurobehavioral regulation. A recent consensus position has been developed across many laboratories that have studied wavelength regulation of the physio- logical and behavioral effects of light in humans and other mammalian species, for the best methods to measure and report light relative to its biological and behavioral effects. The pub- lished consensus provides a free web-based toolbox that allows the calculation of effective irradiance for human melanopsin as well as the opsins in each of the cone and rod photo- receptors that contribute to the physiological effects of light (Lucas et al., 2014). This consensus position on measuring and reporting light stimuli has been formally promulgated by the CIE (Commission Internationale de l’Eclairage, 2015). Row Light source CCT (K) % Blue* Luminous flux (lm) Scotopic content relative to HPS Melanopic content relative to HPS** A PC white LED 2700 17% - 20% 1000 1.77 - 1.82 1.90 - 2.06 B PC white LED 3000 18% - 25% 1000 1.89 - 2.13 2.10 - 2.51 C PC white LED 3500 22% - 27% 1000 2.04 - 2.37 2.34 - 2.97 D PC white LED 4000 27% - 32% 1000 2.10 - 2.65 2.35 - 3.40 E PC white LED 4500 31% - 35% 1000 2.35 - 2.85 2.75 - 3.81 F PC white LED 5000 34% - 39% 1000 2.60 - 2.89 3.18 - 3.74 G PC white LED 5700 39% - 43% 1000 2.77 - 3.31 3.44 - 4.52 H PC white LED 6500 43% - 48% 1000 3.27 - 3.96 4.38 - 5.84 I Narrowband amber LED 1606 0% 1000 0.36 0.12 J Low pressure sodium 1719 0% 1000 0.35 0.10 K PC amber LED 1872 1% 1000 0.70 0.42 L High pressure sodium 1959 9% 1000 0.89 0.86 M High pressure sodium 2041 10% 1000 1.00 1.00 N Incandescent 2851 12% 1000 2.26 2.79 O Halogen 2934 13% 1000 2.28 2.81 P F32T8/830 fluorescent 2940 20% 1000 2.02 2.29 Q Metal halide 3145 24% 1000 2.16 2.56 R F32T8/835 fluorescent 3480 26% 1000 2.37 2.87 S F32T8/841 fluorescent 3969 30% 1000 2.58 3.18 T Metal halide 4002 33% 1000 2.53 3.16 U Metal halide 4041 35% 1000 2.84 3.75 Note: * Percent blue calculated according to LSPDD: Light Spectral Power Distribution Database, http://galileo.graphycs.cegepsherbrooke.qc.CA/app/en/home. The specific calculation, developed for evaluating the potential for affecting sky glow, divides the radiant power contained in the wavelengths between 405 and 530 nm by the total radiant power contained from 380 to 780 nm for each light source. ** Melanopic content calculated according to CIE Irradiance Toolbox, http://files.cie.co.at/784_TN003_Toolbox.xls, 2015, as derived from Lucas et al. (2014). PC = phosphor converted; LED = light-emitting diode. Source: DOE (2016). Table 2. DOE’s selected blue light characteristics of various outdoor lighting sources at equivalent lumen output.

20 LED Roadway Lighting: Impact on Driver Sleep Health and Alertness In developing Table 2, the DOE provided the melanopic content because it is a primary indi- cator of the relative potential for the listed light sources to stimulate human biological responses that are the subjects of much of the AMA’s 2016 statement. Note, however, that influences from other photoreceptors, such as the rods and cones, are known to contribute to biological responses such as circadian and neurophysiological regulation, but in ways that are not fully clear to the medical research community (DOE, 2016). The DOE report also cautioned that the values provided in Table 2 provide only “an idea of the relative potential to cause human health impacts, rather than the actual (if any) impact of the melanopic content” (DOE, 2016). This points to a deficit in the biomedical literature. When that DOE document was distributed in 2016, there were no direct, empirical measurements of the capacity of street and area lighting to evoke circadian, neuroendocrine, or neuro behavioral responses in healthy humans under naturalistic outdoor conditions. This literature review confirms that this is still the case as of August 2018. Gaps in Research Based on the literature review, two major gaps in knowledge have been identified concerning the effects of LED roadway lighting on driver sleep health and alertness. First is a lack of research on the effects of LED roadway lighting on driver alertness. Second is the absence of metrics to quantify an LED light source in terms of its effects on sleep health and alertness. Effects of Roadway Light Sources and Levels on Driver Health and Alertness There is some evidence that light with a high blue content (such as light from LEDs) can increase alertness and enhance cognitive performance in humans (Chellappa et al., 2011; Lehrl et al., 2007). Thus, LED roadway lighting with high blue content might make road users (drivers, pedestrians, and others) more alert, thereby enhancing nighttime traffic safety. To design effec- tive roadway lighting, there is a growing need to understand the relationships between roadway light level, melatonin suppression, and driver alertness and health. However, these relationships have never been investigated in realistic roadway environments. Results from studies conducted in realistic roadway environments could be readily transferred to the real world, as the condi- tions and the dosages encountered would be nearly identical. Such a study would generate rec- ommendations for roadway lighting that are more valid and applicable than those derived from research conducted in indoor laboratories. Further, to inform roadway lighting standards that will reduce unintended negative effects of roadway lighting and improve driver alertness, it is critical to understand the effects of both spectral power distribution and light levels on human health and alertness. Quantifying the LED Light Sources for Human Health and Alertness As previously discussed, there are no direct methods to measure empirically the capacity of LED roadway lighting to evoke circadian, neuroendocrine, or neurobehavioral responses in healthy humans under realistic roadway conditions. Currently, metrics such as color tempera- ture are being used, but these have drawbacks. Metrics such as melanopic lux, melanopic to photopic ratio, and circadian stimulus are currently proposed, but their effectiveness in eliciting a human circadian health and alertness response are yet to be reported in realistic roadway light- ing situations. Identifying such a metric can help state departments of transportation classify LED luminaires so that those with minimal negative health effects and which promote alertness could be installed on roadways.

Literature Review 21   Addressing Gaps in the State of Knowledge Based on these identified gaps, the current study aims to address the following research questions: 1. What are the effects of the illuminance, duration, and spectral power distribution (SPD) of LED roadway lighting on the sleep health and alertness of drivers? How do these effects compare to those of HPS lighting and the absence of roadway lighting? 2. How can the unintended negative consequences of LED roadway lighting (if any) on driver sleep health and alertness be mitigated? 3. If LED roadway lighting affects driver sleep health and alertness, then what metrics for quantifying an LED light source correlate best with eliciting such a response? Project Objective The goal of this study was to determine the impact of LED roadway lighting on driver sleep health and alertness. The research team evaluated the effects of different lighting types and their corneal illuminances on measures of drivers’ sleep health and alertness. Sleep health was evaluated based on melatonin levels in plasma and saliva. Melatonin is a hormone secreted by the pineal gland that affects the human circadian clock system and sleep physiology during the hours of darkness. Sufficiently bright exposure to light can suppress melatonin; thus, melatonin levels can serve as a separate quantitative index related to sleep health. Driver alertness was measured objectively using driver reaction time, percentage of time that a driver’s eyelids are closed over a certain period (PERCLOS), and standard deviation of lateral position (SDLP) as well as subjectively using the Karolinska Sleepiness Scale (KSS).

Next: Chapter 3 - Methods »
LED Roadway Lighting: Impact on Driver Sleep Health and Alertness Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Light emitting diode (LED) technology has revolutionized the lighting industry. The dimming and instant-on capabilities of these light sources along with their high efficiency have allowed lighting designers to overcome some of the limitations of previous technologies, particularly in roadway lighting environments. However, concerns related to the health and environmental impacts of LEDs have been raised.

The TRB National Cooperative Highway Research Program's NCHRP Research Report 968: LED Roadway Lighting: Impact on Driver Sleep Health and Alertness seeks to determine the impact of LED roadway lighting on driver sleep health and alertness.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!