Nocturnal Melatonin Suppression by Adolescents and Adults for Different Levels, Spectra, and Durations of Light Exposure

Abstract

The human circadian system is primarily regulated by the 24-h LD cycle incident on the retina, and nocturnal melatonin suppression is a primary outcome measure for characterizing the biological clock's response to those light exposures. A limited amount of data related to the combined effects of light level, spectrum, and exposure duration on nocturnal melatonin suppression has impeded the development of circadian-effective lighting recommendations and light-treatment methods. The study's primary goal was to measure nocturnal melatonin suppression for a wide range of light levels (40 to 1000 lux), 2 white light spectra (2700 K and 6500 K), and an extended range of nighttime light exposure durations (0.5 to 3.0 h). The study's second purpose was to examine whether differences existed between adolescents' and adults' circadian sensitivity to these lighting characteristics. The third purpose was to provide an estimate of the absolute threshold for the impact of light on acute melatonin suppression. Eighteen adolescents (age range, 13 to 18 years) and 23 adults (age range, 24 to 55 years) participated in the study. Results showed significant main effects of light level, spectrum, and exposure duration on melatonin suppression. Moreover, the data also showed that the relative suppressing effect of light on melatonin diminishes with increasing exposure duration for both age groups and both spectra. The present results do not corroborate our hypothesis that adolescents exhibit greater circadian sensitivity to short-wavelength radiation compared with adults. As for threshold, it takes longer to observe significant melatonin suppression at lower CS levels than at higher CS levels. Dose-response curves (amount and duration) for both white-light spectra and both age groups can guide lighting recommendations when considering circadian-effective light in applications such as offices, schools, residences, and healthcare facilities.

Keywords: age-related change; and spectrum; circadian phase; circadian phototransduction; circadian rhythms; exposure duration; light level; nocturnal melatonin suppression; white light.

Figure 1.

The spectral power distributions of the rated 2700 K and 6500 K LED white light sources used in the study.

Figure 2.

The target CL_A and CS values for the study’s lighting conditions for a 1-h exposure, predicted using the Rea et al. model of circadian phototransduction (Rea et al., 2005; Rea et al., 2012; Rea and Figueiro, 2018).

Figure 3.

Components of the desktop luminaire that was custom-built by the LRC for use in this study (top): (1) satellite link controller, (2) light diffuser, (3) LED linear accent, (4) plywood housing back, (5) pine board base, (6) ½-in. × 1-in., PVC 90° angle (×2), (7) connector cable, (8) installed endcap, and (9) touchpad interface. A prototype of the assembled luminaire is shown in operation (middle) and the typical viewing geometry experienced by the participants is shown at the bottom. The luminaire was placed on a 12-in. high supporting stand that was positioned behind the participants’ personal electronic devices.

Figure 4.

The study protocol. Participants arrived in the laboratory and were held in dim light (< 5 lux at the eye) until the first saliva sample was obtained at 23:00 h. After the first saliva sample was obtained, the desktop luminaires were turned on and 6 additional saliva samples were collected at 30-min intervals.

Figure 5.

Mean salivary melatonin levels for participants in the study’s 2 phases and 2 age groups at each exposure duration. The light levels for Phase 1 (16 adolescents, 16 adults) correspond to the target CS levels of 0.3 and 0.5 for the 2 spectra, and the low light levels for Phase 2 (17 adolescents, 16 adults) correspond to the target CS levels of 0.07 and 0.14 for the 2 spectra. The error bars represent SEM.

Figure 6.

Logistic plot comparing the measured 1-h response data from the present experiment with the predicted responses from the original CS model to validate its accuracy. The data points correspond to mean nocturnal melatonin suppression following 1-h exposures to the 2700 K and the 6500 K sources at different corneal light levels (CL_A, Equation 1). The continuous solid line represents the respective predicted CS values according to Equation 2 and depicts the functional relationship between CL_A and CS plotted without regard for the measured melatonin suppression values obtained in the present study. Goodness of fit (R²) for the logistic function is 0.90. The error bars represent SEM.

Figure 7.

Logistic best-fit (R² = 0.86) to the melatonin suppression data further plotted as a function of photopic illuminance levels, which is a widely used metric to characterize many human visual responses. For each target CS level, the higher photopic light levels for the 2700 K source consistently resulted in lower melatonin suppression compared to the circadian stimulus matched 6500 K source providing lower photopic light levels. The error bars represent SEM.

Figure 8.

Logistic best-fit to the melatonin suppression data as a function of CL_A resulted in an improved fit (R² = 0.96) while rectifying the discontinuity revealed by the fit as a function of photopic illuminance levels (see Figure 7). The error bars represent SEM. It is important to note that while the original fit (see Figure 6) has been derived from the original Rea et al. CS model to validate its accuracy, the logistic best-fits as a function of the photopic illuminance levels (see Figure 7) and CL_A, shown here, are based upon the formulation used by Zeitzer et al. (2000) and adapted from Rea et al., 2012 (Equation 4) to compare the effectiveness of the 2 metrics in characterizing the impact of light on nocturnal melatonin suppression (Rea et al., 2005; Rea et al., 2012).

Figure 9.

The significant main effect of CS level. The asterisks represent p < 0.05 and the error bars represent SEM.

Figure 10.

The significant main effect of exposure duration. The asterisks represent p < 0.05 and the error bars represent SEM.

Figure 11.

The significant interaction between exposure duration and target CS level (p < 0.001). Points marked with an asterisk represent the earliest juncture at which melatonin suppression was significantly > 10% (indicated by dashed line). The error bars represent SEM.

Figure 12.

Using the least squares method, a 4-parameter logistic function (Rea et al., 2012) that converts CL_A to CS, was used to best-fit the melatonin suppression data at each hourly exposure duration, by age group and light source. The warm sources include the 2700 K source from the present study (solid polygons) and a similar 2700 K source from a previous white light study (hollow polygons) (Nagare et al., 2018b). The cool sources include the 6500 K source from the present study (solid polygons) and a similar 5600 K source from the same, previous white light study (hollow polygons) (Nagare et al., 2018b). The error bars represent SEM. The threshold (CS= 0.1) and half-saturation CL_A levels, derived using the respective best-fit plots, were subsequently converted to photopic light levels for the respective spectra and summarized in Table 4.