The effects of spectrum, irradiance, and duration of exposure on light-induced melatonin suppression in healthy adults.

The aim of the study was to test the applicability of the action spectrum of melatonin suppression (MS) based on monochromatic experimental data from Brainard et al. (Journal of Neuroscience, 2001, 21, 6405) and Thapan et al. (Journal of Physiology, 2001, 535, 261) to quantify MS using white light and to characterize its temporal course during the first hour of exposure. Using a within-subject design, 32 healthy adults (12 males/20 females) with free pupil adaptation were exposed to white light of 1904-6996 K in correlated color temperature (CCT), and to blue light (BL, 474 nm). Two "circadian effective" irradiances (Ec = 0.264 W m-2 and 0.528 W m-2) based on Gall's metric (Licht, 2002, 54, 1292) were applied for 60 min using ganzfeld devices. Melatonin concentrations were measured in blood plasma. For 25 participants with complete data sets at Ec = 0.264 W m-2, the r-ANOVA indicated a significant difference in MS between the five different light spectra, F(4,96) = 4.97, p = 0.011. The Bonferroni-corrected post hoc paired t-test (α = 0.005) showed significantly lower melatonin reduction by BL than white light of 2883 K and 1904 K, but not vs. 6997 K and 4663 K. This was due to an unexpectedly high proportion of 11 subjects (=44%) who did not respond to BL with MS ("BL non-responders"). For 14 subjects ("BL responders", =56%), r-ANOVA confirmed equivalent values of MS under all five lighting conditions at both irradiances. Additivity of spectral contributions to the circadian effectiveness of white light is limited, as is reciprocity only with respect to cold white light.