Open Access
Add to BibSonomy
Abstract
Alzheimer disease and related dementias affect 15–20% of elderly people, and 60–70% of these suffer from sleep disturbances. Studies suggest that lighting can improve sleep. The key challenge is how to deliver light effectively. We have designed a lighting system that adjusts spectrum and irradiance on a 24-hour timetable to provide spatially uniform, shadow-free white light with CRI>85 and up to 1000 Lux for day vision and amber light for night vision. To aid sleep, melanopic illuminance varies over 3 orders of magnitude to enable strong suppression of melatonin in the morning/early afternoon, moderate suppression in the evening, and no suppression at night.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Developing appropriate and effective interventions to address dementia-related problems is a critical health care priority. Difficulty falling asleep, fragmentation, wandering, and waking early are highly prevalent (60–70%) [1–3] in persons living with dementia (PWD), and are associated with excessive daytime sleepiness and depression [4]. PWD spend large portions of the day asleep and ∼40% of the night awake, getting out of bed on ∼14 times/night [5], giving rise to safety concerns as PWD are at increased risk for falls when wandering at night [6]. Sleep disturbances, due to their burden to family caregivers are a leading cause for institutionalization [7,8]. In addition, Wang [9] has reported the existence of a feedback loop: poor sleep interferes with brain repair mechanisms progressing dementia resulting in poorer sleep. While there is a lack of evidence to help guide interventions [10], treatment for PWDs falls into three broad categories: 1) pharmacological; 2) sensory stimulation (non-light), and 3) light therapy. The first is largely ineffective or has serious side effects [10]. Sensory stimulation appears to have no effect on sleep [11]. In Hjetland’s 2020 review [12], he reported that while bright light therapy (BLT) is discussed in over 2083 papers, only 31 clinical studies have been conducted: 60% reported improved sleep quality; 40% found no effect. Inter-study comparison is difficult since dosage, timing, and spectral power distribution (SPD) are usually not clearly defined.
Sleep patterns are controlled by intrinsic circadian clocks. These cell-autonomous ‘clocks’ are synchronized by the suprachiasmatic nucleus (SCN), a master pacemaker that uses several neural and endocrine outputs, such as the hormone melatonin, to communicate with peripheral organs. The strongest factor in ‘setting’ the SCN is the natural day-night cycle. The human eye responds to light from ∼390nm to 700nm, with different retinal cells responding to different wavelengths or colors of light. Cones contribute to color (day) vision and rods provide night (grayscale) vision. The recently discovered 5th sensor, photosensitive retinal ganglion cells (ipRGC) [13], despite their small numbers, play a crucial role in many non-visual functions [14]. ipRGC send information regarding light intensity at λ∼485nm directly to the bilateral SCN. Under bright sunlight (ample λ∼485nm light) melatonin production is suppressed. Under conditions of darkness (absence of λ∼485nm light), melatonin production increases in the pineal gland inducing sleepiness. While solar light intensity peaks around this wavelength, as this wavelength is not needed for vision, conventional lighting has limited output at this wavelength leading to insufficient melatonin suppression resulting in daytime sleepiness. Increased napping in the day results in less efficient sleep at night. Nighttime illumination suppresses melatonin production leading to increased wakefulness. This picture is supported by Prayag’s 2019 work [15] which re-analyzed previously conflicting results on the effects of irradiance by scaling them by the sensitivity of ipRGC, defining a new parameter – melanopic illuminance (units mLux) – to complement the commonly used photopic illuminance (which weights the irradiance by the sensitivity of the cones). Results [15] (Fig. 2) clearly show the scaled data follows a single S-curve (log scale) demonstrating the irradiance of light weighted by the sensitivity of ipRGC determines melatonin suppression. There is both an initiation threshold (∼1.5 mLux, 10% suppression) and a saturation level (305 mLux, 90% suppression) for melatonin suppression. The need to include the melanopic response in specifications of lighting is now recognized, details agreed upon, and standards published [16].
In the last few years, two clinical trials using Dynamic Lighting (DL), lighting in which intensity and SPD are varied (over at least part of the day), reported that varying lighting can reduce sleep problems of PWDs [17,18]. Both studies were conducted (1) in residential care settings, (2) with participants required to sit under specially designed lighting equipment in a (3) controlled environment. The first led by E. van Lieshout-van Dal [17], varied the contribution of cool white (CCT=6500K) and warm white (CCT=2700K) luminaires along with the total illuminance so that the maximum intensities and the bluest light was found from 10am to 6pm (max 1900 Lux) dropping in the evenings to 385 Lux and lower CCT. They observed that the frequency of night-time bed wanderings decreased from 11 to 5 times per night. In the second (Dec 2019), M Figueiro [18] simultaneously changed the correlated color temperature (CCT) and irradiance to obtain similar results.
Unfortunately, requiring a PWD to wear a helmet, sit in a bright lightbox for a few hours per day or even stay in one place, seems to be wishful thinking. In fact, ensuring compliance is a common concern. The current methodology cannot generally be applied for PWDs nor a 24-hour lighting program be easily implemented. In addition, due to the way light is delivered, it is difficult to determine overall exposure. Indeed, the authors of both clinical studies, Figueiro [18] and van Lieshout-van Dal [17] state that the biggest challenge is “to find a practical method for effectively delivering the lighting intervention to the eyes of people with dementia”.
The purpose of this paper is to present a practical method for “effectively delivering the lighting intervention to the eyes of people with dementia” in such a way that compliance is ensured and the requirements for day and night vision are met.
2. System design
Figure 1 provides a schematic of installation of lights in the ∼22 m2 rooms (circumference ∼20 m, modeled on a typical long-term care room). The upper inset is a picture of the room in Taiwan taken from beside the bed while the lower inset is a picture of the Ressam Gardens Model Suite (McMaster Innovation Park in Hamilton, ON). Hidden in the coves (Taiwan) or crown molding (Ressam Gardens), extending around the room’s perimeter [19]), parallel LED strips are mounted on the hypotenuse of a 3-4-5 triangle. This design serves to hide the LED sources from the PWD. On the upper side of the cove (not visible) is a glossy Al sheet which serves as both a high reflectivity specular reflector and heat sink. The room’s matte white ceiling acts as diffuse reflector to scatter uniform indirect light into the room below. The angle at which the strips are placed ensures that emitted light’s first reflection is off the ceiling rather than a wall. Matte light green walls scatter the light that reaches them preferentially in the horizontal plane contributing to the vertical/horizontal uniformity of the lighting environment. The baseboard trim hides a single strip of waterproof amber LEDs (IP68). High gloss white paint specularly reflects the light onto the wood floor which acts as a diffuse reflector to provide sufficient light for night vision while minimizing sleep disruption [6]
Fig. 1. Schematic of room layout with an upper cove (Ice Blue, Warm White, and RGB LED strips) and a baseboard (Amber LED strip). Both cove and baseboard extend around the room. (Insets) Photos of demo room in Taiwan (upper) and Ressam Gardens Model Suite at McMaster Innovation Park (lower)
In total, the four sets of 5m strips, warm white (CCT∼2700 K), ice blue, and RGB (120 LEDs/m) in the cove provide day lighting, amber in the baseboard trim (60 LED/m) provides night lighting, contain 8400 LEDs. Standard LED strips rather than specialized strips are used as certified strips are readily available in most markets. Two constant voltage 500 W Power Supplies (MeanWell) independently provide a maximum of I=20A of output at VDC=24V each.
Lighting is controlled by a single micro-controller (Espressif Systems, ESP-WROOM-32) programmed in C on a 24-hr cycle which dims individual LED strips (pulse width modulation, PWM, at 8kHz) allowing both total irradiance and the SPD to be controlled. Battery operated 3-way switches (externally identical to large button wall switches commonly used in the home environment), allow the PWD to turn lights ON or OFF with the SPD and irradiance appropriate for that time of day or night. Wireless control allows these switches to be placed where most convenient for the PWD (e.g. by the door (Fig. 1 (top inset)), beside the bed). An additional, separate emergency push button, placed near the door, allows a caregiver to override the 24-hr lighting protocol to provide high intensity light (CCT ∼6000K) in the case of emergency. For safety of the PWD, a Passive Infrared (PIR) Motion Sensor automatically turns on the lights (at the level appropriate for the time of day or night) whenever motion is detected as PWDs often neglect to turn on lights. All communication is done using ESPNOW() protocol (Expressif Systems). The key design criteria: make control of the system appear identical to that which the PWD has experienced in their life.
3. Lighting
Figure 2(a) presents the spectrum of the various LEDs used in the system in comparison with the solar spectrum. For reference, the peaks of the α-opic action spectra for the S-cone, M-cone, L-cone, rods (Rhodopic), and ipRGC (Melanopic) are denoted by arrows. None of these LEDs optimally, and only red and amber do not, excite the ipRGC.
Fig. 2. (a) SPD of RGB, Ice Blue (IB), Warm White (WW), and Amber LEDs compared with the solar spectrum. Arrows indicate the peaks of the α-opic action spectra for the short (SC) medium (MC) and long (LC) cones, rods (R) and ipRGC (M) [20]. (b) White light SPD at selected times. The Melanopic (Rhodopic) action spectra is shown in yellow (gray) [20].
However, by varying the absolute and relative intensities of these LEDs strips, both the photopic illuminance and melanopic illuminance along with their ratio can be varied. The bulk of the light in the room is provided by the warm white and ice blue strips during the daytime and the amber strip during the night. Ice blue has been chosen for use in this system due to its significant spectral component around λ=485nm. The main function of the RGB strip is to allow the addition of sufficient amounts of Red-Blue-Green so that the light remains “white” with low values of Δuv and high values of Color Rendering Index (CRI).
Figure 2(b) presents the SPD and Table 1 the corresponding α-opic parameters [20] at selected times during the day and night recorded at eye level (1.5m) formed by combining the output of the LED strips. Irradiance and illuminance are the mean of 15 measurements taken on a 1m square grid. SPD is independent of position. In Fig. 2(b) both the melanopic and the rhodopic action spectra (night vision) are shown in the background in yellow and gray respectively. The overlap between the SPD and the melanopic action spectra increases from the early morning rising to a maximum at 12:30 (noon) and then gradually decreases throughout the late afternoon (15:00) and into the early and late evenings. During the night, only the amber strip is active in order to provide sufficient light to stimulate the rods for night vision (overlap between the rhodopic action spectra (gray) and the amber emission) while limiting the overlap with the melanopic action spectra. Amber (λc=600nm, FWHM=12nm) is uniquely suited to perform both duties. The relative effectiveness of a SPD on the stimulation of ipRGC at a given photopic illuminance is given by the ratio of melanopic/photopic (M/P) lux (Table 1, last column). As seen in the table, the value varies by an order of magnitude during the 24-hour cycle with a minimum during night (M/P=0.15) and a maximum M/P=1.45 in the early afternoon. During the late evening hours an M/P=0.54(∼⅓ of that at noon) allows for adequate lighting for visual purposes while reducing melatonin suppression.
Along with the SPD of the luminaries, the irradiance is also varied over the 24-hour period. Figure 3(a) presents the photopic Lux (black) in the vertical direction over the 24-hour cycle. It is designed to be like the day/night cycle of natural sunlight, howbeit with an extended evening. To enable adequate day vision, photopic illuminance gradually increases starting at 6:30am reaching a maximum of over 1000 Lux from noon to early afternoon. Intensity then gradually falls, first to 500 Lux at 6pm and then further declining in intensity (after dinner) in the early evening to reach a minimum of 300 Lux. Superimposed on Fig. 3(a) with black dot-dashed lines are key illuminance levels from the literature. Early research indicated that 1000 Lux is required for setting the circadian clock [21] while the 500 Lux and 300 Lux are the minimum illumination levels specified by the Illuminating Engineering Society of North America’s Recommended practice for Lighting and the Visual Environment for Seniors and the Low Vision Population (ANSI/IES RP28-18) [16] for common areas and bedroom areas respectively. 1.5 Lux is the minimum illuminance for night vision [16]. As seen in the figure, the light intensity exceeds that required for both day and night vision.
Fig. 3. (a) Photopic (black) and melanopic (blue) illuminance over 24 hours in the vertical and horizontal directions respectively. Gray horizontal lines indicate recommended minimum photopic illuminance for setting the biological clock [21], common areas, bedroom and at night [16]. Blue horizontal lines indicate the mean (dash) melanopic illuminance for 10%, 50%, 90% melatonin suppression [15] and the range (solid) for 50% suppression in individuals [22]. (b) Corresponding CCT (black) and CRI (blue) of the white light. The gray (blue) horizontal line indicates the maximum (minimum) value of CCT (CRI) for comfortable vision in seniors [16]
For vision, the white lights color temperature and color rendering are crucial. Figure 3(b) presents both the CCT (black) and CRI (blue) over the 24 hour cycle. Superimposed are two horizontal lines, a blue line indicating the minimum acceptable CRI for seniors (CRI=85) [16] and a black line indicating the threshold for discomfort among seniors (CCT=6500K) [16]. The implemented lighting system maintains CRI>85 throughout the day with values being >90 for most of the daylight hours indicating the white light has good color rendering properties. The CCT of the white light increases during the morning, reaches a maximum of 6300 K in the early afternoon, and then gradually decreases to ∼2500 K at 21:30 before entering night mode. The varying CCT and intensity provide stimulus to the visual system as to the time of day.
Additional factors such as glare, uniformity, and shadows are also crucial parameters for senior vision [16] and fall avoidance [23]. (∼40 percent of adults over 65 years fall each year. Falls are the 2nd leading cause of accidental death in the population overall [23]). Shadows [24] lead to falls as a PWD mistakes them for a discontinuity on the floor to be stepped around. Table 2 compares the quality and uniformity of the lighting distribution with a commercial direct (diffuse) state-of-the-art lighting system installed in the same room. While both systems were glare-free, the present lighting system is also completely shadow-free. Illuminance is slightly higher and much more uniform than the direct lighting system with only 15% (18%) variation across the room in the vertical (horizontal) direction compared with 29% (25%) for the commercial system. In addition, as the PWD gaze shifts from the horizontal plane towards the ceiling, the PWD will only see an 80% increase in intensity while with the direct system the intensity will not only more than double but is also highly dependent on whether the PWD is looking at the light fixtures or not. While the wall-plug efficiency (η) is considerably lower than the direct system (∼23 lumens(vertical)/W), the difference between the two systems is less in the more important horizontal direction. In addition, this number can be improved with high diffuse reflectivity paint on the ceiling and higher efficiency LED strips.
Table 2. Uniformity of illuminance within the rooms compared with a commercial direct diffuse lighting
Figure 3(a) also presents the horizontal (56% of the vertical) melanopic lux over the 24-hour cycle (blue) as this corresponds to gaze angles of PWDs and the experimental conditions of Brainard’s original measurements [25] that were reprocessed 11 years later [15]. Dashed blue lines indicates the illuminance that results in a 10% (threshold), 50%, and 90% suppression of melatonin [15]. Phillips [22] observed considerable variation between individuals with values for 50% suppression ranging from 3.4 to 200 mLux based on a 5-hour exposure at constant illuminance (horizontal gaze angle). (Note: Phillips reported values of the Lux. The author has converted these values to mLux using their light source’s SPD). These lines are also shown. Melanopic illuminance follows the photopic illuminance throughout the day and night being modulated by changes in the SPD induced M/P ratio which declines starting in the early afternoon (Table 1). For average individuals, the intensity is sufficient to provide for >90% melatonin suppression during regular working hours (8 am to 5 pm) reaching nearly 100% in the early afternoon. During the evening the suppression drops to ∼70%. During the night, the amber lighting (0.41 mLux), is well below the threshold for melatonin suppression. Based on Phillips’ work, the high intensity in the early afternoon is sufficient to suppress melatonin by 90% in even the least sensitive individuals and the light levels in the late evening will not suppress melatonin in even the most sensitive individuals. (It is crucial to take into consideration individuals rather than just the mean for healthy individuals as it is likely to be individuals with sensitivities differing from the mean that will encounter sleep disorders. In addition, PWD are generally seniors whose sensitivity to light declines with age [16]). In summary, by varying the intensity of photopic illuminance by two order of magnitude (10 to 1000 Lux) and varying the M/P ratio (by changes to the SPD) by an order of magnitude (M/P=0.15 to 1.45) a three order of magnitude variation in the melanopic illuminance is achieved.
4. Discussion and conclusions
The total energy consumption of the system (wall-plug) as implemented is <0.4 kW hr/m2/day. The cost of electricity is US$0.05/m2 (Ontario, Canada) or $28/month for a 20 m2 room (assuming lighting is on 24 hours per day). We have recently replaced the current LED strips with more energy-efficient ones reducing power consumption, and electricity costs to US$18/month. Obviously, in a communal living environment, the per person operating costs would be less as one would expect the PWD not to spend 24 hours isolated in their room but rather to gather in common areas at least for meals.
Due to the substantial order of magnitude variation in the sensitivity of individuals to light [22], applying this lighting system in an individual home should be proceeded by first determining first the individual’s regular (pre-dementia) day/night cycle and then correlating the degree of melatonin suppression with melanopic illuminance. This information can then be used to tailor both the timing and delivered illuminance in the controlling C program to the PWD’s needs. The lighting system developed here has the flexibility to meet the needs of the most sensitive and least sensitive individuals.
The durability and acceptability of the lighting system has been run tested in the two rooms (model suite for Ressam Gardens Retirement Home at McMaster Innovation Park, Canada, and the demo room at Yuan Ze University in Taiwan) from February of 2020 continuing to the present – a total of 20 months. The response to the system has been uniformly favorable with those spending time in the room stating that they feel the indirect lighting is more comfortable and relaxing than the commercial system and those visiting the model suite commenting favorably on the lighting in comparison to the commercial system (details to be presented elsewhere). The only issue that we have found is a slight blue-shift over a period of a year due to slightly less efficient conversion of blue light by the phosphors in the Ice blue and a reduction in output of the red emission in the RGB strips. This can be compensated for in software by slightly modifying the duty cycles of the LEDs.
In summary, a 24-hour aesthetically appealing indirect dynamic white lighting system has been designed and tested to provide light to meet the visual and non-visual requirements of elderly patients. For day vision, it delivers uniform shadow-free white light with high CRI, a high ratio of horizontal to vertical illuminance of sufficient illuminance at a CCT<6500 K. For circadian purposes, the three orders of magnitude variation in melanopic illuminance provides for a high degree of melatonin suppression during the morning to early afternoon, with suppression declining during the late afternoon and evening. During the night, amber lighting provides sufficient stimulation of the rods for vision without stimulating the ipRGC cells. It is our dream that just as a positive feedback loop exists between increasing levels of Alzheimer dementia (defined as the accumulation of amyloid-β and tau) and increasing sleep fragmentation [9], dynamic white lighting will result in improved sleep resulting in reduced levels of Alzheimer dementia resulting in improved sleep – essentially reversing this feedback loop. Future advances in LED technology (e.g., organolead halide perovskite nanocrystals [26]) should allow even more control over white light illumination.
Funding
Ministry of Science and Technology, Taiwan (108-2221-E-155-050, 109-2221-E-155-051).
Acknowledgments
We thank Deb Bryson and Mary Burnett of the Alzheimer Society Foundation Brant, Haldimand Norfolk, Hamilton Halton for their valuable input and feedback regarding the lighting design. JDW acknowledges his mother’s (Marjorie Ruth White) 7 plus year war with dementia in opening his eyes to the importance of God-provided white light in setting the circadian rhythm.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. B. Guarnieri, F. Adorni, M. Musicco, I. Appollonio, E. Bonanni, P. Caffarra, C. Caltagirone, G. Cerroni, L. Concari, F. I. I. Cosentino, S. Ferrara, S. Fermi, R. Ferri, G. Gelosa, G. Lombardi, D. Mazzei, S. Mearelli, E. Morrone, L. Murri, F. M. Nobili, S. Passero, R. Perri, R. Rocchi, P. Sucapane, G. Tognoni, S. Zabberoni, and S. Sorbi, “Prevalence of sleep disturbances in mild cognitive impairment and dementing disorders: A multicenter Italian clinical cross-sectional study on 431 patients,” Dement Geriatr Cogn Disord 33(1), 50–58 (2012). [CrossRef]
2. A. Rongve, B. F. Boeve, and D. Aarsland, “Frequency and Correlates of Caregiver-Reported Sleep Disturbances in a Sample of Persons with Early Dementia,” Journal of the American Geriatrics Society 58(3), 480–486 (2010). [CrossRef]
3. M. Moran, C. A. Lynch, C. Walsh, R. Coen, D. Coakley, and B. A. Lawlor, “Sleep disturbance in mild to moderate Alzheimer’s disease,” Sleep Medicine 6(4), 347–352 (2005). [CrossRef]
4. D. K. Y. Miu and S. S. L. Szeto, “Sleep disturbances among a group of dementia participants,” Journal of Clinical Gerontology and Geriatrics 3(3), 105–109 (2012). [CrossRef]
5. E. Bonanni, M. Maestri, G. Tognoni, M. Fabbrini, B. Nucciarone, M. L. Manca, S. Gori, A. Iudice, and L. Murri, “Daytime sleepiness in mild and moderate Alzheimer’s disease and its relationship with cognitive impairment,” J Sleep Res 14(3), 311–317 (2005). [CrossRef]
6. N. Hanford and M. Figueiro, “Light Therapy and Alzheimer’s Disease and Related Dementia: Past, Present, and Future,” JAD 33(4), 913–922 (2013). [CrossRef]
7. A. Kesselring, “Emotional and physical demands on caregivers in home care to the elderly in Switzerland and their relationship to nursing home admission,” The European Journal of Public Health 11(3), 267–273 (2001). [CrossRef]
8. J. E. Gaugler, A. B. Edwards, E. E. Femia, S. H. Zarit, M.-A. P. Stephens, A. Townsend, and R. Greene, “Predictors of Institutionalization of Cognitively Impaired Elders: Family Help and the Timing of Placement,” The Journals of Gerontology Series B: Psychological Sciences and Social Sciences 55(4), P247–P255 (2000). [CrossRef]
9. C. Wang and D. M. Holtzman, “Bidirectional relationship between sleep and Alzheimer’s disease: role of amyloid, tau, and other factors,” Neuropsychopharmacology 45(1), 104–120 (2020). [CrossRef]
10. J. McCleery and A. L. Sharpley, “Pharmacotherapies for sleep disturbances in dementia,” Cochrane Database of Systematic Reviews 2020(11), CD009178 (2020). [CrossRef]
11. T.-D. Dimitriou and M. Tsolaki, “Evaluation of the efficacy of randomized controlled trials of sensory stimulation interventions for sleeping disturbances in patients with dementia: a systematic review,” Clin. Interventions Aging 12, 543–548 (2017). [CrossRef]
12. G. J. Hjetland, S. Pallesen, E. Thun, E. Kolberg, I. H. Nordhus, and E. Flo, “Light interventions and sleep, circadian, behavioral, and psychological disturbances in dementia: A systematic review of methods and outcomes,” Sleep Medicine Reviews 52, 101310 (2020). [CrossRef]
13. D. M. Berson, “Phototransduction by Retinal Ganglion Cells That Set the Circadian Clock,” Science 295(5557), 1070–1073 (2002). [CrossRef]
14. D. M. Dacey, H.-W. Liao, B. B. Peterson, F. R. Robinson, V. C. Smith, J. Pokorny, K.-W. Yau, and P. D. Gamlin, “Melanopsin-expressing ganglion cells in primate retina signal colour and irradiance and project to the LGN,” Nature 433(7027), 749–754 (2005). [CrossRef]
15. A. S. Prayag, R. P. Najjar, and C. Gronfier, “Melatonin suppression is exquisitely sensitive to light and primarily driven by melanopsin in humans,” J Pineal Res 66(4), e12562 (2019). [CrossRef]
16. The IES Lighting for the Aged and Partially Sighted Committee, Recommended Practice for Lighting and the Visual Environment for Seniors and the Low Vision Population (ANSI/IES RP28-18), (Illuminating Engineering Society of North America (IESNA) (2016).
17. E. van Lieshout-van Dal, L. Snaphaan, and I. Bongers, “Biodynamic lighting effects on the sleep pattern of people with dementia,” Build. Sci. 150, 245–253 (2019). [CrossRef]
18. M. G. Figueiro, B. Plitnick, C. Roohan, L. Sahin, M. Kalsher, and M. S. Rea, “Effects of a Tailored Lighting Intervention on Sleep Quality, Rest–Activity, Mood, and Behavior in Older Adults With Alzheimer Disease and Related Dementias: A Randomized Clinical Trial,” Journal of Clinical Sleep Medicine 15(12), 1757–1767 (2019). [CrossRef]
19. A. Assefa, P.-J. Chen, X. L. Ho, and J. D. White, “On the potential of solid state LED strips utilizing an organic color converter for non-line of sight visible light communication,” Opt. Express 25(20), 24242–24250 (2017). [CrossRef]
20. CIE, CIE S 026/E:2018 CIE System for Metrology of Optical Radiation for IpRGC-Influenced Responses to Light, CIE (2020).
21. R. F. Riemersma-van der Lek, “Effect of Bright Light and Melatonin on Cognitive and Noncognitive Function in Elderly Residents of Group Care Facilities,” JAMA 299(22), 2642–2655 (2008). [CrossRef]
22. A. J. K. Phillips, P. Vidafar, A. C. Burns, E. M. McGlashan, C. Anderson, S. M. W. Rajaratnam, S. W. Lockley, and S. W. Cain, “High sensitivity and interindividual variability in the response of the human circadian system to evening light,” Proc. Natl. Acad. Sci. U. S. A. 116, 12019–12024 (2019). [CrossRef]
23. L. N. Saftari and O.-S. Kwon, “Ageing vision and falls: a review,” J Physiol Anthropol 37(1), 11–14 (2018). [CrossRef]
24. T. Philpot, “No shadows,” Nursing Standard 20(23), 24–25 (2006). [CrossRef]
25. G. C. Brainard, D. Sliney, J. P. Hanifin, G. Glickman, B. Byrne, J. M. Greeson, S. Jasser, E. Gerner, and M. D. Rollag, “Sensitivity of the Human Circadian System to Short-Wavelength (420-nm) Light,” J. Biol. Rhythms 23(5), 379–386 (2008). [CrossRef]
26. G.C. Adhikari, P.A. Vargas, H. Zhu, A. Grigoriev, and P. Zhu, “Tetradic phosphor white light with variable CCT and superlative CRI through organolead halide perovskite nanocrystals,” Nanoscale Adv. 1(5), 1791–1798 (2019). [CrossRef]
